article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the fitzhugh - nagumo system is a simplified model for nerve conduction which has become a paradigm for dynamical and neuronal investigations since , despite of being considerably simple , it is still able to reproduce qualitatively the main dynamical behaviors of realistic neural models .
the literature on this subject is quite vast and , in many times , rather confusing with respect to definitions and names .
we adopt here the nomenclature and the main conventions of @xcite . the fitzhugh - nagumo model for a single externally excited neuron
is given by the following two dimensional non - autonomous system of first order differential equations [ fhn ] \
{ l v = f(v ) - w + ( t ) + w = v - w . where @xmath0 and @xmath1 are non - negative parameters and @xmath2 stands for the external excitation .
the usual and simplest choice for the function @xmath3 is f(v ) = v - , and we also adopt it hereafter .
our approach , however , can be easily modified to deal with any other kind of function @xmath3 .
typically , the @xmath1 parameter in the fitzhugh - nagumo system is considered to be small , implying the usual dynamical decomposition in fast @xmath4 and slow @xmath5 modes @xcite . in the absence of the external excitation ( @xmath6 )
, the intersection of the nullclines @xmath7 and @xmath8 determines the possible fixed points for the system .
since we are mainly interested in the simplest possible chaotic solutions for the system ( [ fhn ] ) , we will restrict ourselves to the cases exhibiting only one fixed point , which corresponds to the values of @xmath9 .
the three - dimensional phase space of the non - linear system ( [ fhn ] ) is known to accommodate a rich dynamical behavior , ranging from periodical bursting to chaotic oscillations .
the limit @xmath10 is called a singular perturbation of ( [ fhn ] ) since we are , effectively , reducing the phase space dimension of the system ( see , for instance , @xcite ) .
in such a limit , which preserves many dynamical properties of the original system , the variable @xmath11 is constrained to move along the nullcline @xmath12 , implying that the dynamics of the system are such that [ constraint ] w(t)=f(v(t ) ) + ( t ) , for all @xmath13 . by substituting ( [ constraint ] ) in the second equation of ( [ fhn ] )
, we have [ eq ] ( f(v ) + ( t ) ) = v- ( f(v ) + ( t ) ) , which is a first order non - autonomous differential equation with , consequently , a two - dimensional phase space .
equation ( [ eq ] ) is not separable for generic @xmath2 and , hence , there is no hope of obtaining its general solutions . nevertheless ,
since it effectively corresponds to a two - dimensional system , one might ask at this point if the phase space of the singularly perturbed fitzhugh - nagumo equation ( [ eq ] ) would not be classified according to the poincar - bendixson theorem , what would leave no room for any chaotic dynamical behavior .
the answer is definitely no .
equation ( [ eq ] ) can be written as [ eq2 ] ( 1-v^2 ) = ( 1-)v- ( t)+ ( - ( t ) ) , from where we see clearly that the system is singular on the vertical ( on the plane @xmath14 ) lines @xmath15 .
thus , we can not use the poincar - bendixson theorem , nor its recent generalization for piecewise lipschitz continuous systems @xcite , globally on the phase space of ( [ eq2 ] ) . as we will see , equation ( [ eq2 ] ) does indeed exhibit chaotic behavior , even for its simplest forms .
there is a well known and studied particular case for which we do have analytical solutions for ( [ eq ] ) .
it corresponds to a neuron excited by periodic instantaneous ( rectangular ) pulses of the form [ pulse ] ( t ) = \ { ll 0 , & for 0 t < , + a , & for t < t , . with @xmath16 , @xmath17 , and non - negative @xmath18 .
this kind of external excitation is physiologically reasonable and has been considered before in numerous analysis of the fitzhugh - nagumo system , see , for instance , @xcite and the references cited therein . in particular , we will follow here the approach of @xcite and introduce the one - dimensional map @xmath19 corresponding to a stroboscopic map@xcite of the equation ( [ eq ] ) sampled on periods @xmath20 , _
i.e. _ , if @xmath21 , then @xmath22 .
this kind of map can be used to explore many aspects of the underlying dynamics of the system . for a comprehensive recent review on maps in neuronal dynamics ,
see @xcite . in the next section ,
we show how to determine numerically and accurately the map @xmath23 together its lyapunov exponent , which will be used eventually to investigate the main point of the present work : the chaotic oscillations of ( [ eq ] ) under an external excitation of the type ( [ pulse ] ) .
the stroboscopic map @xmath23 for the equation ( [ eq ] ) requires the evaluation of the solutions @xmath24 at @xmath25 , for generic values of @xmath26 .
we do not adopt here any further simplification as , for instance , those ones considered in @xcite corresponding to assume a piecewise linear
shaped @xmath3 function , which makes equation ( [ eq ] ) linear and , consequently , simplifies considerably the problem .
far from the discontinuities of @xmath2 , equation ( [ eq2 ] ) reads [ eq1 ] ( 1-v^2 ) = v^3 + ( 1-)v - a , for @xmath27 , whereas for @xmath28 we have exactly the same expression , but with @xmath29 .
equation ( [ eq1 ] ) can be easily integrated by a simple quadrature , leading to the following general solution [ int ] _
v_0^v ds = t - t_0 .
in order to obtain the matching conditions for @xmath24 on the discontinuities of @xmath2 , let us integrate ( [ eq ] ) in the distributional sense , for instance , from @xmath30 to @xmath31 , and then take the limit @xmath32 .
one gets [ jump1 ] f(v(^+ ) ) = f(v(^- ) ) - a. analogously , we will have [ jump2 ] f(v(0^+ ) ) = f(v(0 ^ - ) ) + a. the dynamical picture in the phase space of ( [ eq ] ) is as follows .
far from the singularities of @xmath2 , the solution is constrained to move on the surface ( [ constraint ] ) .
when a pulse appears ( for instance , for @xmath33 ) , the solution jumps instantaneously ( same @xmath13 ) and horizontally ( same @xmath34 ) to a new value of @xmath11 according to the cubic equation ( [ jump1 ] ) , which in fact corresponds to jump horizontally from the curve on the plane @xmath14 given by ( [ constraint ] ) with @xmath6 to the new one with @xmath35 .
this situation is well illustrated , for the case of piecewise linear
shaped @xmath3 functions , in @xcite and @xcite .
the present situation is completely analogous to those ones .
there are others discontinuities in the dynamics of @xmath11 .
as it was already noticed , an inspection of ( [ eq1 ] ) reveals that @xmath36 diverges for @xmath15 . assuming a @xmath0 parameter small enough to assure that the right handed side of ( [ eq1 ] ) has , for a given value of @xmath18 , only one zero and that it is located in the interval @xmath37
, we will have @xmath38 for @xmath39 ( respectively , @xmath40 for @xmath41 ) .
a solution starting , say , at @xmath39 is driven towards the point @xmath42 .
on this point , since @xmath36 diverges on it , the solution jumps instantaneously and horizontally to the point @xmath43 , and then it is driven towards @xmath44 and so on , giving origin to a periodic oscillation in the well known fashion .
notice that , inside the interval @xmath37 , the directions are reversed , _
i.e. _ , @xmath40 for @xmath45 ( respectively , @xmath38 for @xmath46 ) , where @xmath47 stands for the fixed point of ( [ eq1 ] ) , which is unique by hypothesis . in this way , a solution starting at @xmath45 will be driven towards @xmath42 and start a new periodic solution , jumping to @xmath43 and so on .
solutions starting in the regions @xmath48 will be driven towards the singularity @xmath49 and also start the periodic oscillations .
one can , without loss of generality , remove the regions @xmath50 and @xmath48 from our map , since no oscillating solution is ever injected there . for generic @xmath51 and @xmath52 , the integral ( [ int ] ) can be evaluated by expanding the rational fraction in term of partial fractions .
( see the appendix a for these details , for the numerical scheme , and for the algorithm . )
however , for @xmath53 ( the van der pol limit of the fitzhugh - nagumo equation ) , the problem is greatly simplified , despite of keeping unaltered its main dynamical properties . in such a limit
, we have simply [ vdp ] - = t - t_0 .
in fact , the main dynamical properties of the model are independent of the details of the function @xmath3 , provided it has the usual and characteristic n " shape . the solution for a fitzhugh - nagumo equation with a generic @xmath3 under excitations of the type ( [ pulse ] )
is given by [ int1 ] _
v_0^v ds = t - t_0 .
the matching and the jumping conditions for @xmath24 are essentially the same .
our stroboscopic map can be easily modified to incorporate the general quadrature ( [ int1 ] ) , which might be , if necessary , evaluated numerically with good accuracy .
we have set up a numerical code @xcite , using scilab , that takes into account the jumps @xmath54 and the match conditions ( [ jump1 ] ) and ( [ jump2 ] ) in the general solutions ( [ int ] ) .
the pertinent details are presented in the appendix a. we have used it to construct accurately the stroboscopic map @xmath55 for any values of the parameters @xmath0 , @xmath18 , @xmath56 , and @xmath20 .
figure [ fig1 ] depicts a typical map for @xmath57 .
this particular situation , of course , corresponds to a regular periodic solution ( no external excitation ) . provided the value of @xmath0
be such that there is only one fixed point and it is located in the interval @xmath37 , the dynamical behavior seems to be essentially independent of @xmath0 .
since we have removed the region @xmath37 of our analysis , we redefine our intervals in order to have @xmath58\cup[1,2]\to [ -1,1]$ ] by simple shifts .
the obtained maps are typically piecewise smooth and the interval @xmath59 $ ] can be effectively partitioned into several dynamically disjoint and inequivalent sets .
since we are dealing with piecewise smooth one - dimensional maps @xmath60 , we could in principle study their dynamical properties from their lyapunov exponent @xmath61 a positive lyapunov exponent ( [ lamba ] ) would correspond undoubtedly to chaotic oscillations in this context .
since our maps are piecewise smooth , it is indeed possible that the interval @xmath62 $ ] be partitioned into regions with different lyapunov exponent for the same map .
the index @xmath63 in ( [ lamba ] ) is a reminder of this important fact .
the map depicted in fig .
[ fig1 ] , for instance , should have a vanishing lyapunov exponent for all the interval @xmath62 $ ] , and we could verify it numerically with very good accuracy .
this map , in particular , corresponds to a periodic solution with period @xmath64 , but sampled with period @xmath65 . since @xmath66 is irrational , we do not expect either fixed or periodic points of any order , _
i.e. _ , there should exist no @xmath67 such that @xmath68 . since the underlying solution is regular and periodic ,
we expect neither positive lyapunov exponents for this situation . however , by varying the map parameters we can easily induce the apparition of fixed ( @xmath69 ) and periodic points .
if a fixed point is stable @xmath70 , its basin of attraction @xmath71 will have a lyapunov exponent @xmath72 on the other hand , unstable fixed points @xmath73 are typically good candidates to induce chaotic dynamics .
neutral fixed points @xmath74 could also in principle give origin to chaotic behavior .
we will return to this point in the last section .
figure [ fig2 ] depicts the stroboscopic map for the case with @xmath75 , @xmath76 , @xmath77 , and @xmath78 , which clearly exhibits three fixed points , which are shown separately in figure [ fig3 ] , with some cobweb iterations .
the first fixed point ( ordered according to the value of @xmath63 ) is clearly a stable one . for any point @xmath63 of its basin of attraction @xmath79
, we will have @xmath80 .
the dynamical behavior is rather clear .
the solutions starting in @xmath79 at @xmath81 , after undergoing a transient regime ( the oscillations far from the fixed point @xmath82 ) , will tend to @xmath82 , _
i.e. _ , they will tend towards a periodic behavior with period @xmath20 .
the forcing term @xmath2 is clearly driving the solution in this case , a rather common behavior for damped excited oscillators .
much more interesting , however , are the other two fixed points .
they are clearly unstable @xmath73 and the solutions tend to depart from each other , see the cobweb iterations depicted in fig .
. however , the map is such that the trajectory is constantly re - injected in the region with @xmath83 , leading eventually to a positive lyapunov exponent for this region of the interval @xmath84 . for this case in particular , one has @xmath85 .
a map with @xmath83 implies exponential deviations for initially nearby trajectories , leading in this way to the most distinctive characteristic of chaotic dynamics : the extreme sensitivity to initial conditions .
our numerical set up allows also the investigation of the origin of the unstable fixed points in the stroboscopic map and , consequently , the onset of the chaotic behavior in the fitzhugh - nagumo system , as the system parameters vary .
an example is shown in figure [ fig4 ] .
we keep @xmath86 , @xmath87 , and @xmath78 fixed , and vary @xmath18 .
the onset of chaos corresponds to the appearance of the unstable fixed points , which occurs for @xmath88 , the second curve , from top to bottom , in figure [ fig4 ] . by increasing the value of @xmath18 ,
the distance between the two fixed points increases , enlarging in this way the chaotic region , which corresponds essentially to the region bounded by the two unstable fixed points . the corresponding lyapunov exponent @xmath89 also increases , indicating a stronger chaotic behavior
however , for @xmath90 , the map becomes discontinuous in this region , the branches corresponding to the two fixed points separate from each other and a gap between them appears .
the trajectories that were typically chaotic now tend to a limit cycle in this gap , and the system becomes rather abruptly a regular one with a vanishing lyapunov exponent .
the last curve in figure [ fig4 ] illustrates namely this situation .
the overall dynamical picture is the following . for @xmath91 , the system is regular , _
i.e. _ , there is no chaotic motion , there is no sub - interval with positive lyapunov exponent . for this case ,
typically , we have subintervals with vanishing and negative lyapunov exponents .
the former corresponds to those situations for which the external excitation is not strong enough to disturb considerably the natural periodic solutions of the fitzhugh - nagumo system . on the other hand
, the latter corresponds to the appearance of stable fixed points , indicating that the external excitation could effectively be driving the system . for @xmath92
, we have new dynamical behavior associated to the regions with positive lyapunov exponents , namely the regions limited by the unstable fixed points .
we still have , however , coexisting regions with vanishing and negative lyapunov exponents . for @xmath93 , we yet have the two unstable fixed points with the solutions departing from each other , but due to the gap between them , the solutions are not anymore continuously re - injected in regions with @xmath94 .
the system ceases to exhibit chaotic motion in this case .
it is also noticeable that the basin of attraction of the stable fixed points grows in size for increasing values of @xmath18 .
this is hardly a surprise since , with larger amplitudes , the external excitation tends to take over the dynamics of the system more easily .
such an overall behavior seems to be quite generic and robust against small variations in @xmath0 , @xmath56 , and @xmath20 , and the shape of @xmath3 as well .
we have also detected a property of the system which dynamical origin is still unclear for us : smaller values of @xmath56 seems to favor the appearance of unstable fixed point leading to chaos in the fitzhugh - nagumo system .
moreover , the route to chaos associated with the the parameter @xmath56 ( and the period @xmath20 ) seems to be essentially the same associated to @xmath18 .
figure ( [ fig5 ] ) depicts the situation of the unstable fixed points for different values of @xmath56 .
the situation is very similar to that one depicted in ( [ fig4 ] ) .
the routes to chaos , in particular , are complementary in the sense that both routes are associated to deformations in the stroboscopic map around the unstable fixed point .
variations in @xmath18 imply deformation mainly along the left branch of the map ( fig .
( [ fig4 ] ) ) , while variations in @xmath56 do it along the right branch ( fig .
( [ fig5 ] ) ) . these results and scenarios are ,
again , quite robust against variations in @xmath0 , provided they do not change the number of the fixed points in the original fitzhugh - nagumo system ( [ fhn ] ) .
we have studied the singularly perturbed limit of periodically excited two - dimensional fitzhugh - nagumo systems by introducing a stroboscopic one - dimensional map sampled with the period of the external rectangular pulse excitation .
the pertinent numerical and computational details are presented in the appendix a. by exploring the lyapunov exponent of the stroboscopic map , we look for chaotic behavior in the original system .
we show how to identify and characterize the chaotic regions from the analysis of unstable fixed points of the stroboscopic map .
our results corroborates and , in fact , complement those ones presented in @xcite , where the same analysis is employed , but for the rather common approximation of considering a piecewise linear
shaped @xmath3 function in the fitz - hugh - nagumo equation .
the main emphasis in @xcite was the characterization of the chaotic dynamics from the bifurcation structure of the stroboscopic map .
our approach could also be used for this kind of analysis .
figure [ fig6 ] , for instance , depicts the region containing the unstable fixed points for the second- and forth - iterated stroboscopic map of the figure [ fig2 ] ( @xmath75 , @xmath76 , @xmath95 , and @xmath78 ) .
one can see clearly the presence of unstable periodic points , another typical feature of chaotic maps .
our method can be also applied to the systems with different functions @xmath3 and , in particular , to that one considered in @xcite .
in fact , since they are linear , the general solution corresponding to ( [ int ] ) is trivial , but a complex dynamics do indeed arise due to the jumping and matching conditions .
our algorithm presented in the appendix can be easily adapted for this situation , allowing the analysis of the lyapunov exponents of the cases considered in @xcite , for instance .
we finish by remarking that it is indeed possible to have chaotic dynamics for maps with neutral @xmath96 fixed points .
this is the case , for instance , of the so - called intermittent systems .
the usual lyapunov exponent vanishes for these systems , but they can still exhibit sub - exponential instabilities characterized by a positive subexponential lyapunov exponent " @xmath97 with @xmath98 , see @xcite for further details . the parameter @xmath99 depends on the details of the map in the vicinity of the neutral fixed point .
we have identified some stroboscopic maps with fixed points which might be neutral .
the possibility that these neutral fixed points could generate sub - exponential instabilities with positive @xmath100 for the fitzhugh - nagumo system is now under investigation .
the authors are grateful to fapesp and cnpq for the financial support . as wishes to thank professor leon brenig for the warm hospitality at the universit libre de bruxelles , where part of this work was carried on .
the stroboscopic map and its lyapunov exponent are constructed numerically from the general solution ( [ int ] ) and the matching condition for @xmath24 on @xmath15 and on the discontinuities of @xmath2 . for @xmath86 , the general solution ( [ int ] )
is given explicitly by the simple expression ( [ vdp ] ) . in order to write the solution for the general case @xmath101 , let us consider the cubic polynomial [ poly ] p_3(v ) = v^3 + 3v - 3a , with @xmath102 .
since we have @xmath103 , the polynomial ( [ poly ] ) has only one real root .
it is already in the so - called depressed form , so it is rather simple to obtain its real root @xmath47 ( the fixed point ) by elementary algebra . by using the standard vieta s substitution ,
one has [ a2 ] v _ * = w _ * - , with w_*^3 = a + . notice that for @xmath29 , we have @xmath104 , leading to @xmath105 .
we remind our hypothesis that @xmath0 must be small enough ( and hence @xmath106 must be large enough ) to assure that @xmath107 for a given value of @xmath18 . the polynomial ( [ poly ] )
can be factorized as @xmath108 , where the quadratic polynomial p_2(v ) = v^2 + v_*v + 3+v_*^2 , has no real solution , by construction .
finally , the rational function in the integrand of ( [ int ] ) can be expanded as [ frac ] = ( + ) , where @xmath109 , @xmath110 , and @xmath111 are the unique solution of the linear system [ a3 ] ( ccc 1 & 1 & 0 + v _ * & -v _ * & 1 + 3+v_*^2 & 0 & -v _ * ) ( c a_1 + a_2 + a_3 ) = ( c -1 + 0 + 1 ) .
the case @xmath29 @xmath112 corresponds to @xmath113 , @xmath114 , and @xmath115 . by using ( [ frac ] ) ,
the integral in ( [ int ] ) can be evaluated with elementary methods , leading finally to h(v ) - h(v_0 ) =
t - t_0 , where h(v ) = .
( notice that the case corresponding to @xmath75 is given by the simpler expression ( [ vdp ] ) . ) with the explicit function @xmath116 , one can determine the stroboscopic map by following the description of the dynamics of section 2 .
this procedure is summarized in the algorithm 1 .
the lyapunov exponent @xmath89 can be evaluated from its definition for one - dimensional maps ( [ lamba ] ) .
the derivative @xmath117 can be easily evaluated numerically .
there is , however , a potential problem in the numerical evaluation of @xmath117 .
since our map is piecewise smooth and we do not know a priori the location of its discontinuities , the left and right approximations for the derivative do not always converge to the same value .
in fact , we identify the discontinuities by monitoring the sudden variations in one of them . if the left and right approximation for the derivative are close , we take the simple average as the approximation . if they differ by orders of magnitude ( indeed , a factor of @xmath118 was enough in our computations ) , we discarding the largest ( in modulus ) and keep the remaining one as the approximation for the derivative .
@xmath119 evaluate @xmath47 , @xmath109 , @xmath110 , and @xmath111 for @xmath29 @xmath120 evaluate @xmath121 such that @xmath122 evaluate @xmath123 such that @xmath124 evaluate @xmath47 , @xmath109 , @xmath110 , and @xmath111 for the input value of @xmath18 @xmath125 @xmath126 @xmath120 evaluate @xmath121 such that @xmath127 evaluate @xmath123 such that @xmath128 return @xmath123 | we consider the singularly perturbed limit of periodically excited two - dimensional fitzhugh - nagumo systems .
we show that the dynamics of such systems are essentially governed by an one - dimensional map and present a numerical scheme to accurately compute it together with its lyapunov exponent .
we then investigate the occurrence of chaos by varying the parameters of the system , with especial emphasis on the simplest possible chaotic oscillations .
our results corroborate and complement some recent works on bifurcations and routes to chaos in certain particular cases corresponding to piecewise linear fitzhugh - nagumo - like systems .
chaos , excitable systems , fitzhugh - nagumo systems , neuron model |
99 g. marchesini et al . ,
_ comput . phys
. commun . _ * 67 * ( 1992 ) 465 .
g. corcella and m.h .
seymour , hep - ph/9908335 .
d collaboration , b. abbott et al .
lett . _ * 80 * ( 1998 ) 5498 .
cdf collaboration , t. affolder et al .
, fermilab - pub-99/220-e . | we show vector boson transverse momentum distributions at the tevatron , obtained by running the herwig monte carlo event generator with matrix - element corrections .
we compare our results with some recent d and cdf data .
ur1589 + er/40685/939 + hep - ph/9911536 + november 1999 vector boson production at hadron colliders is a fundamental process to test quantum chromodynamics and the standard model of the electroweak interactions .
the lowest order processes @xmath0 and @xmath1 are not sufficient to perform reliable phenomenological predictions , but the initial - state radiation has to be taken into account . a possible way to deal with such multiple emissions consists in running a monte carlo event generator .
standard monte carlo algorithms @xcite describe parton cascades in the soft / collinear approximation , with ` dead zones ' in the phase space which can be filled by the using of the exact first - order matrix element .
in @xcite we implemented matrix - element corrections to the herwig simulation of drell yan interactions : we filled the missing phase space using the exact @xmath2 matrix element ( hard corrections ) and corrected the shower in the already - populated region using the exact amplitude for every hardest - so - far emission ( soft corrections ) . for @xmath3 production at the tevatron , about 4% of the events
are generated in the dead zone , about half of which are @xmath4 events .
similar results hold for @xmath5 production as well . an interesting observable
to study is the vector boson transverse momentum @xmath6 , which is the object of many theoretical and experimental analyses .
while in the parton shower approximation it has to be @xmath7 , after matrix - element corrections a fraction of events with larger values of @xmath6 is to be expected . in fig .
[ wqt ] ( a ) , we plot the @xmath3 @xmath6 spectrum at the tevatron according to herwig 5.9 , the latest public version , and to herwig 6.1 , the new version including matrix - element corrections to drell yan processes .
herwig has the option to vary the intrinsic transverse momentum @xmath8 of the incoming partons , which is set to zero , its default value , in the distributions shown in fig .
[ wqt ] ( a ) .
we observe a remarkable impact of the corrections : after some @xmath6 the 5.9 version does not give events anymore , while herwig 6.1 still has some events generated via the exact matrix element . in fig .
[ wqt ] ( b ) we compare some recent d data @xcite on @xmath3 production at the tevatron with the herwig 6.1 results , which we corrected in order to take the detector smearing effects into account .
we find good agreement overall ; we also consider the option of @xmath9 gev , but do not see any relevant effect after the detector corrections . as far as @xmath5 production is concerned ,
we have some preliminary cdf data @xcite , already corrected for detector effects , which we compare with herwig 6.1 in fig .
[ zqt ] , where the options @xmath8=0 , 1 and 2 gev are investigated .
the agreement is acceptable and the role of the implemented matrix - element corrections is crucial in order to succeed in fitting in with the data for @xmath10 gev . at very low @xmath6 ,
the best fit is obtained by setting @xmath11 gev . in fig
. [ ratio ] , we plot the ratio of the @xmath3 and the @xmath5 transverse momentum spectra , both normalized to unity , for different values of @xmath8 .
although it can be seen from fig .
[ zqt ] ( b ) that the @xmath5 @xmath6 spectrum depends strongly on @xmath8 at low @xmath6 , the ratio of the @xmath3 and @xmath5 spectra is insensitive to it .
this is good news for the @xmath3 mass measurement in hadron collisions , as this ratio is one of the main theory inputs that is needed .
a strong dependence on unknown non - perturbative parameters like @xmath8 could limit the accuracy of the @xmath3 mass measurement at the tevatron and , ultimately , at the lhc .
we have added matrix - element corrections to herwig s treatment of vector boson production in hadron collisions .
they make an enormous difference at high transverse momentum @xmath6 , but little at low @xmath6 .
although the dependence of the results on the non - perturbative intrinsic @xmath6 of partons in the proton ( @xmath8 ) is quite strong at low @xmath6 , it is very similar in the @xmath3 and @xmath5 cases , so that the ratio of the two @xmath6 spectra is almost independent of @xmath8 . |
euler s 3-body problem is a soluble special case of the general 3-body problem , in which a test particle moves in the gravitational field of two gravitational sources - point masses - which are fixed in space .
it is liouville integrable since , besides the energy and one component of the angular momentum , there is a third constant of the motion ( see e.g. @xcite ; see @xcite for a hyperbolic space version of the problem which is still integrable ) .
a natural relativistic generalisation of this problem is to consider the motion of test particles in a weyl solution describing two schwarzschild black holes , which are kept in equilibrium due to a strut in between them . for this problem
there is no known analogue to euler s constant ( see @xcite for a discussion about this point ) ; equivalently , there is no known separability for the geodesic equations .
generic geodesics have , therefore , to be studied numerically . for
the special case in which the two black holes have equal mass , there is , however , a 1 + 2 dimensional totally geodesic sub - manifold in which geodesics are liouville integrable .
in particular it is straightforward to derive all circular causal orbits .
regular ( on and outside an event horizon ) static black holes in vacuum general relativity do not admit causal circular geodesics below a certain critical radius , which lies outside the event horizon . below this radius ,
the angular momentum term in the radial equation of motion , usually centrifugal ( with respect to the black hole ) , becomes centripetal , making equilibrium impossible .
this change in character of the angular momentum term has a geometric interpretation in terms of optical geometry @xcite : define outwards " as the direction in which fixed points sets of the optical geometry s isometry group ( spheres ) increase their proper size ; then , the angular momentum term is always outwards " .
similar considerations may be made about stationary black holes .
however , in this case the optical geometry is a randers - finsler rather than riemannian geometry , making the interpretation more subtle @xcite . in this paper
we show that the optical geometry of the aforementioned relativistic euler s 3-body problem is richer than the standard one of the schwarzschild black hole . to be concrete let the two black holes in the weyl solution considered have equal mass @xmath2 and separation @xmath3 between them .
this choice of masses implies that there is a discrete @xmath0 symmetry of the solution .
the set of fixed points of this discrete symmetry is a totally geodesic sub - manifold , @xmath11 .
the optical geometry of this sub - manifold then has the following feature : for sufficiently small but non - zero @xmath9 the notion of outwards `` , defined as above , changes continuously twice , rather than just once .
this means that , on @xmath11 , there is an annulus - like region in which ' ' outwards " means a decreasing radial weyl canonical coordinate .
the boundaries of this annular region are the loci of two circular photon orbits , and the region itself is a forbidden band for time - like circular geodesics , which are allowed everywhere outside this band . as @xmath9 increases , the two boundaries of the annular region approach one another and coalesce when @xmath7 , where @xmath8 is the golden ratio ! for larger @xmath9 , @xmath11 admits no circular photon orbits and time - like circular geodesics are allowed everywhere , just as in the @xmath0 symmetric euler s 3-body problem .
considering a weyl solution with an array of @xmath10 , rather than two , black holes along a line , still possessing a @xmath0 discrete symmetry , leads to two cases . for @xmath10 even ( odd ) , we essentially recover the optical geometry of the @xmath12 ( @xmath13 ) case .
we shall also consider charged black holes in einstein - maxwell theory , by using a solution generating technique @xcite that allows the introduction of a charge parameter in weyl solutions .
for black holes , the charge parameter is the ratio of charge to mass , which is the same for all the black holes in the solution . in this case , as we increase the charge for fixed @xmath14 , the proper area of the forbidden band increases ; for @xmath15 , this area diverges as we reach extremality , as a consequence of the infinite ` throat ' developed by extremal black holes . keeping the charge and @xmath2 fixed ,
the area of the forbidden band decreases as @xmath9 is increased , just as for the uncharged case , becoming zero at some maximum value of @xmath9 .
this paper is organised as follows .
after a brief discussion of the newtonian euler s 3-body problem in section 2 , we describe in section 3.1 the weyl solutions that shall be used in the relativistic version of the problem , as well as the circular null and time - like orbits . in section 3.2
the charged case is considered . in section 4
the interpretation of the forbidden band for time - like circular orbits is given in terms of optical geometry .
in section 5 we briefly discuss the case of multiple black holes .
we close with some final remarks .
in euler s 3-body problem , a.k.a . the two - centre kepler problem ( see e.g. @xcite ) , with both masses equal to @xmath16 , the motion of a test particle moving in the symmetry plane between the two masses obeys ( ) ^2=2e - v ( ) , v()=- , [ radequ ] where @xmath17 is a radial cylindrical coordinate , @xmath18 are the energy and angular momentum per unit mass and @xmath9 is the distance from either mass to the symmetry plane .
we shall use geometrised units throughout .
circular orbits are seen by extremising the potential .
they obey = .
this equation has solution for any @xmath17 , given @xmath9 and @xmath16 .
the angular momentum @xmath19 goes to zero ( infinity ) as @xmath20 ( @xmath21 ) - fig .
[ newtonian ] .
thus circular orbits exist for any radial distance in the symmetry plane .
this is therefore the behaviour we expect in the relativistic version of the problem for large @xmath22 , and indeed it is the behaviour we shall find .
the behaviour for small @xmath22 will , however , be quite different . , of circular orbits in the symmetry plane of the two - centre kepler problem , as a function of their angular momentum , @xmath19 .
the two equal masses , which we set to unity , have separation @xmath3 , and various values of @xmath9 have been used .
observe that circular orbits exist for any @xmath17.,height=297 ] in the relativistic problem that shall be considered in the next section , the newtonian gravitational potential of one or multiple _ rods _
( rather than point - like masses ) of mass @xmath16 , infinitesimal width and mass density @xmath23 plays an important role , since it is a key ingredient in the construction of weyl solutions describing one or multiple black holes .
let us remark that , as a newtonian problem , the passage from point particles to two equal rods , aligned with the direction of separation , does not alter the previous conclusion : in the symmetry plane , circular orbits still exist at any radial distance from the symmetry axis .
weyl geometries @xcite in four space - time dimensions admit two commuting , mutually orthogonal killing vector fields . in vacuum , the metric may always be written in the form ds^2=-e^2u(,z)dt^2+e^-2u(,z ) .
[ metric ] the vacuum einstein equations , @xmath24 , reduce to a harmonic equation in an auxiliary euclidean three space , @xmath25 , in cylindrical coordinates @xmath26 , _
^3u=0 , [ udifeq ] and two partial differential equations for @xmath27 _ z k=2_u _ z u , _
k= , [ kdifeq ] which become line integrals once the solution for @xmath28 is known .
a solution with two black holes is obtained by taking @xmath28 to be the potential of two finite rods of zero width and linear density @xmath29 .
the rods are placed at @xmath30 and in the intervals @xmath31 $ ] and @xmath32 $ ] .
then e^2u(,z)= , [ u2]where we have introduced the quantities r_k , _ kz - a_k . introducing further
y_ijr_ir_j+_i_j+^2 , the solution of , given , is e^2k(,z)= .
[ k2 ] an integration constant could be added to @xmath33 .
we have chosen this constant to be zero . with this choice , identifying the azimuthal
coordinate with standard period @xmath34 , the metric is smooth , on the symmetry axis for @xmath35 and @xmath36 . for @xmath30 and @xmath37
there is a conical excess , given by = 2(e^-k(=0,a_2<z < a_3)-1)=2 .
the vacuum solution described by , and is actually a three parameters family of solutions .
physically , the three parameters may be taken to be the two black holes masses and the distance between them .
the black hole masses may be computed as komar integrals : m_1=_s_1d = , m_2=_s_2d = , where @xmath38 is the dual 1-form to the time - like killing vector field @xmath39 and the 2-surfaces @xmath40 correspond to the location of the two finite rods in weyl coordinates .
an asymptotic expansion shows that the adm mass is m_adm = m_1+m_2 . for the distance
we shall take the coordinate distance in weyl coordinates 2l = a_3-a_2 .
this is a monotonic function of the proper distance z=_a_2^a_3dz , and hence a good measure thereof . in order to have a totally geodesic sub - manifold we require @xmath41 .
we also choose a symmetric coordinate system : @xmath42 and @xmath43 .
the proper distance between the two black holes may then be expressed as z = e ( ) , where @xmath44 is a complete elliptic integral of the second kind .
it follows that @xmath45 is a totally geodesic sub - manifold , which has induced metric given by with @xmath45 .
geodesics in this sub - manifold obey @xmath46 where the dot represents derivative with respect to an affine parameter and we have introduced the energy @xmath47 and angular momentum @xmath19 for a particle of mass @xmath48 . from , circular
null geodesics are determined by the extrema of the potential v()==()^4 .
[ potentialv ] for @xmath15 , i.e the single black hole limit , making the coordinate transformation = r , the potential becomes v(r)=(1- ) .
[ potschwarzschild ] this is the effective potential for photons in a schwarzschild black hole and in schwarzschild coordinates .
it has the well known extremum at @xmath5 .
another solution for a constant @xmath49 photon orbit is @xmath6 , corresponding to the null geodesic generator of the horizon .
thus , in weyl coordinates we have two extrema , for @xmath15 , at _ cno_1(l=0)=0 , _ cno_2(l=0)=m_adm . [ extrema]for @xmath50 , @xmath51 has no extrema , since it is approximately @xmath52 .
the interpolation between these two behaviours is as follows - fig [ potential ] .
as @xmath9 grows , the two extrema approach one another , in both coordinate and proper distance , and merge for @xmath53 . beyond this value of @xmath9
, @xmath51 has no extrema . to see the value of @xmath54 ,
observe that the extrema of are given by f(,m_adm , l)= , f(,m_adm , l)- . to solve for @xmath17
it is convenient to introduce @xmath55 and @xmath56 .
then the last equation becomes = - .
[ curve ] take to define a curve @xmath57 , between @xmath58 and @xmath59 ; at these values of @xmath60 , @xmath61 .
these two points " of the curve correspond to @xmath62 and @xmath63 given in .
constant @xmath64 slices of the curve have two solutions for @xmath65 , one solution for @xmath66 and no solution for @xmath67 . to determine @xmath68 , it is convenient to introduce (
1+y_min)^2/3 . then , differentiating , extremising @xmath64 and replacing back into gives the quadratic equation ^2 - 3 + 1=0 , whose only solution leading to a positive @xmath68 is = 1 + , .
@xmath8 is the golden ratio . using the recurrence property of the golden ratio ^n+1=^n+^n-1
, we arrive at @xmath69 , and therefore at m_k = l_max l_max = m_k , , where @xmath70 is the golden ratio conjugate .
thus , quite strikingly , the two null circular orbits merge when the ratio between the komar mass of either black hole and the semi - distance between them is the golden ratio !
this merging happens for weyl radial coordinate = m_k . , which has two extrema at @xmath71 and @xmath72 ; ii ) @xmath73 , for which the two extrema have moved closer to one another ; iii ) @xmath74 , for which the two extrema have merged .
we have fixed @xmath75 in the plots .
right : radial coordinate @xmath17 of the photon circular orbits as a function of @xmath9.,title="fig:",height=297 ] , which has two extrema at @xmath71 and @xmath72 ; ii ) @xmath73 , for which the two extrema have moved closer to one another ; iii ) @xmath74 , for which the two extrema have merged .
we have fixed @xmath75 in the plots .
right : radial coordinate @xmath17 of the photon circular orbits as a function of @xmath9.,title="fig:",height=297 ] ( 0,0)(0,0 ) the significance of the circular photon orbits described in the last subsection can be seen by considering circular time - like orbits . from these
are obtained by extremising the potential v ( ) & = & e^2u(,0)+j^2 + & = & ( ) ^2 + ( ) ^4 , where @xmath19 is now the angular momentum per unit mass .
the extrema of this potential are given by f(,m_adm , l)= .
[ timeuncharged]in fig .
[ rhoj ] we display the solution of eq . for @xmath17 , in terms of @xmath19 , for fixed values of @xmath9 .
one observes that for @xmath76 there is a _ forbidden band _ for time - like circular geodesics , which is the region between the two photon circular orbits .
this behaviour should be contrasted with that exhibited in fig .
[ newtonian ] for the newtonian case and is illustrated in fig .
[ orbits ] . of the circular time - like orbit as a function of the angular momentum per unit mass @xmath19 , for various values of @xmath9 and fixing @xmath75 .
one observes that there is a forbidden region @xmath77 $ ] , for @xmath78 , where @xmath79 are the radii of the two circular null orbits .
the time - like orbits approach the photon orbits for @xmath80 , from either side.,height=336 ] ( 0,0)(0,0 ) , the two circular photon orbits sit at @xmath5 ( in schwarzschild coordinates ) and at the horizon ( top ) . as @xmath9 increases these two photon orbits approach one another ( middle ) and merge as @xmath81 is approached ( bottom ) .
the two photon orbits delimit the forbidden annular region for circular time - like orbits ( coloured region).,height=364 ] ( 0,0)(0,0 ) let us start by recalling that in a reissner - nordstrm black hole , the circular null orbits are found as extrema of the potential v(r)=(1-+ ) , in standard schwarzschild - type coordinates , where @xmath82 is the charge to mass ratio of the black hole . extremising this potential
gives the location of the circular null orbit outside the horizon r_cno_2=+ , whereas the second null orbit , with constant radial coordinate , relevant for our analysis is located at the horizon r_cno_1=m_adm+ .
thus , the forbidden band for circular time - like orbits exists for all black holes in the reissner - nordstrm family . in the extremal case
, this band acquires the special property of having infinite area , since the horizon , at @xmath83 , is at an infinite proper radial distance from any point with radial coordinate @xmath84 , which is the case for @xmath85 . in order to consider two
charged reissner - nordstrm black holes we shall now discuss charged weyl solutions .
the background fields are ds^2=-e^2|u(,z)dt^2+e^-2|u(,z ) , a=-(,z)dt .
[ metriccharged ] the electrovacuum einstein - maxwell equations r_=2(f_f_^ -g_f_f^ ) , d_f^=0 , give rise to the set of equations _ ^3|u = e^-2|u , _
^3=2 , and _
z k=2_|u _ z |u-2e^-2|u__z , _ k =- e^-2|u[(_)^2- ( _ z ) ^2 ] .
[ kdifeq2 ] this set of equations is more involved than eq . - , in particular because there is no longer a linear equation to be solved .
however , if one takes e^2|u(,z)=1-(,z)+(,z)^2 , where @xmath82 is a constant , and ( , z)= , the einstein - maxwell equations for reduce to exactly the eq . - @xcite .
thus we take @xmath28 and @xmath27 to be the same as in the uncharged case , and .
the background will then describe a double reissner - nordstrm solution , wherein both black holes have the same charge to mass ratio @xmath82 .
the study of circular null geodesics will follow that of the uncharged case , _ mutatis mutandis _ ; they will be given by the extrema of the potential v()==^4 , where @xmath86 may be read off from .
such extrema obey f ( , m_adm , l)=- , where q .
[ newcharge ] , in terms of the distance @xmath9 for various values of the charge @xmath87 defined by .,height=336 ] ( 0,0)(0,0 ) the result for the radii of the photon circular orbits in terms of the distance @xmath9 for various values of the charge is shown in fig .
[ chargedplot ] , where we used the coordinate @xmath88 .
the first feature we wish to emphasise is that , as @xmath9 is increased for fixed charge , the two photon orbits approach one another , just as for the uncharged case , and coalesce for a maximum value of @xmath9 .
thus , the forbidden band for time - like circular orbits exists for all possible values of the charge , for sufficiently small @xmath9 .
the second feature we want to mention is that , fixing @xmath89 and increasing the charge , the forbidden band has larger area . for @xmath15
this area diverges as extremality is approached , as mentioned above . in fig .
[ areas ] we exhibit an example with @xmath90 . in the extremal case ,
for which the solution is completely regular on and outside the event horizon , the analysis may also be done using the majumdar - papapetrou @xcite form of the solution : ds^2=-+h(,z)^2(d^2+^2d^2+dz^2 ) , a=- , where h(,z)=1++ . in this extremal case ,
circular null orbits exist in the @xmath45 sub - manifold as long as the equation ( ^2+l^2)^3/2=2m_k(^2-l^2 ) , has solutions for real and positive @xmath17 , which for fixed @xmath2 is possible when @xmath9 obeys ll_max()^3/2m_k . as a function of @xmath17 , for fixed @xmath75 , @xmath91 , and two values of the charge .
the vertical lines correspond to the radii of the null circular orbits , and the shaded region corresponds to the area of the forbidden band for time - like circular orbits divided by @xmath92 . in the @xmath93 case
the exterior circular photon orbit is outside the @xmath17 range .
observe that the area increases with the charge.,height=336 ] ( 0,0)(0,0 )
a @xmath94 dimensional static geometry with @xmath95 isometry group may always be expressed by the line element ds^2=g_tt(r)dt^2+g_rr(r)dr^2+g_(r)d_d-2 , [ general ] where @xmath96 is the line element on the @xmath97-sphere .
the optical geometry is the effective spatial geometry seen by light rays : ds^2_optical = dr^2+d_d-2 , defined where @xmath98 . for the spacetime ,
the radial equation of motion of a particle with mass , energy and angular momentum @xmath99 , respectively is orbits are therefore determined by the extrema of the potential v(r)= , which is both the coefficient of the angular momentum term in and the inverse of the proper radius squared of the sphere line element in the optical geometry .
this potential pushes the test particle in the direction of increasing proper size of the spheres in the optical geometry .
we dub this direction as _
outwards_. if it coincides with our nave notion of `` outwards '' , the angular momentum term may still be interpreted as a _
term , as in flat space .
time - like circular orbits are only possible in a region where this term and the mass term in originate forces in opposite directions . for a single schwarzschild black hole of mass @xmath16 ,
circular time - like orbits are possible for @xmath100 in schwarzschild coordinates .
thus they are forbidden , outside the horizon , for @xmath101 . in this case
, the optical geometry is ds^2_optical=+d_2 . the proper size ( area ) of the 2-sphere
is therefore @xmath102 , where the potential @xmath103 is given by .
the area increases with @xmath49 for @xmath104 , but decreases for @xmath101 - fig .
[ optical1 ] ( left ) .
thus the centrifugal force " points towards the black hole , for @xmath101 .
since the mass term is always attractive towards the black hole , there can be no circular time - like orbits in the region @xmath101 .
( an annulus - like region , when @xmath105 ) . in the single schwarzschild case ,
the optical metric is singular at @xmath106 ; for @xmath107 , the proper size is negative since @xmath108 . in the double - schwarzschild case
, the annulus - like region lies in between the two extrema of the curves .
we have set the individual black hole masses to unity.,title="fig:",height=288 ] ( an annulus - like region , when @xmath105 ) . in the single schwarzschild case ,
the optical metric is singular at @xmath106 ; for @xmath107 , the proper size is negative since @xmath108 . in the double - schwarzschild case
, the annulus - like region lies in between the two extrema of the curves .
we have set the individual black hole masses to unity.,title="fig:",height=288 ] ( 0,0)(0,0 ) for the double schwarzschild solution the optical geometry in the symmetry plane is ds^2_optical = e^-4u(,0 ) .
[ metricoptical ] the proper size ( perimeter ) of the @xmath109 circle is therefore given by @xmath110 , where the potential @xmath51 is given by .
the forbidden band for circular time - like orbits is the region wherein the radial coordinate decreases outwards " , and therefore , the centrifugal force " becomes directed to the symmetry axis - fig .
[ optical1 ] ( right ) - preventing again the existence of time - like circular orbits .
we shall now consider a weyl solution with @xmath10 uncharged black holes . then is replaced by e^2u(,z)= .
[ u2many]this function @xmath28 describes the newtonian potential of @xmath10 rods of infinitesimal width and mass density @xmath29 , located at @xmath30 and with @xmath111 coordinate in the intervals , respectively , , , , .
=- a_2n+1-n , n=1
2n , the geometry still admits a discrete @xmath0 symmetry , of which @xmath45 is a fixed point set .
then , the circular null orbits in this totally geodesic sub - manifold are obtained by extremising the potential v()== ( ) ^2 , which yields the condition _ n=1 ^ 2n=1 .
we set a_2n - a_2n-1=2m_k , a_2n+1-a_2n=2l , n=1
n , such that the coordinate distance between each two black holes is @xmath3 and their komar mass is @xmath2 . in fig .
[ multiple ] we have plotted the radial coordinate of circular null orbits versus @xmath9 , for various numbers of black holes . for @xmath10 even , the behaviour is quite similar to that discussed for @xmath12 .
moreover _ n+ l_max= m_k . for @xmath10 odd , the behaviour is quite similar to that of a single schwarzschild black hole , due to the existence of a black hole in the symmetry plane .
black holes as a function of @xmath9 .
left ( right ) : n odd ( even).,title="fig:",height=288 ] black holes as a function of @xmath9 .
left ( right ) : n odd ( even).,title="fig:",height=288 ] ( 0,0)(0,0 )
in this paper we have studied a relativistic version of euler s 3-body problem : the motion of test particles in a @xmath0 invariant weyl solution describing two ( or more ) black holes .
we have considered both uncharged and charged black holes , within einstein - maxwell theory .
generically , these solutions have conical singularities on the symmetry axis . in the extremal case , however , they become a special case of the majumdar - papapetrou spacetimes , and are completely regular on and outside the event horizon @xcite . for sufficiently small distance between the two black holes , there are two circular null geodesics in the symmetry plane corresponding to the set of fixed points of the @xmath0 symmetry .
the interior one is stable ( against perturbations in the symmetry plane ) , while the exterior one is unstable .
these orbits delimit a forbidden region for circular time - like geodesics . in the optical geometry
, the proper size of the @xmath112 isometry group orbits grows towards the symmetry axis , rather than away from it , in this region .
this unusual property is correlated with a physical effect that may be , heuristically , described as the centrifugal force becoming directed towards the symmetry axis .
thus , it can not balance the gravitational attraction , preventing the existence of time - like circular orbits .
the existence of a stable circular null geodesic is associated with a behaviour which may be described as an ` optical fibre ' or ` wave guide ' like geometry .
photons with sufficiently low energy will be caught in a potential well in the radial direction , as may be observed in fig .
[ potential ] . thus
, as they move in the @xmath109 direction , photons will oscillate between a maximum and a minimum radii .
the wave guide like geometry may allow light waves to travel around the @xmath109 direction without decreasing their amplitude significantly .
as the distance between the two black holes is increased , the two null circular orbits approach one another , eventually coalescing . for the uncharged case this happens
when the ratio between the komar mass of each black hole and the semi - distance between the black holes is the golden ratio .
one may ask whether the geodesic equations in the full spacetime , rather than just the symmetry plane , are liouville integrable . in @xcite
a similar question was studied for extremal black holes in einstein - maxwell - dilaton theory , for arbitrary dilaton coupling .
these authors showed that the geodesic motion is generically chaotic .
there is only a very special case of integrable motion : when the dilaton coupling is that of kaluza - klein theory .
for the generic case studied herein , of charged weyl solutions that may be continuously connected to the extremal black holes of einstein - maxwell theory ( with zero dilaton coupling therefore ) , it seems quite unlikely that the motion may be integrable .
observe however , that the uncharged double schwarzschild solution may also be continuously connected to the extremal black holes of kaluza - klein theory .
thus , it would be interesting to understand where , along this trajectory in the space of solutions , integrability appears .
we are very grateful to g. w. gibbons and e. teo for correspondence and comments on drafts of this paper .
is supported by a junior research grant from centro de fsica do porto .
is supported by a cincia 2007 research contract .
this work as been further supported by the fct grant cern / fp/83508/2008 .
g. w. gibbons and c. m. warnick , `` hidden symmetry of hyperbolic monopole motion , '' j. geom .
* 57 * ( 2007 ) 2286 [ arxiv : hep - th/0609051 ] . c. m. will , `` carter - like constants of the motion in newtonian gravity and electrodynamics , '' phys .
* 102 * ( 2009 ) 061101 [ arxiv:0812.0110 [ gr - qc ] ] .
g. w. gibbons , c. a. r. herdeiro , c. m. warnick and m. c. werner , `` stationary metrics and optical zermelo - randers - finsler geometry , '' phys
. rev .
d * 79 * ( 2009 ) 044022 [ arxiv:0811.2877 [ gr - qc ] ] .
j. v. jos and e. j. saletan , `` classical dynamics ( a contemporary approach ) , '' cambridge university press , 1998 . j. b. hartle and s. w. hawking , `` solutions of the einstein - maxwell equations with many black holes , '' commun . math .
* 26 * ( 1972 ) 87 .
g. w. gibbons , `` non - existence of equilibrium configurations of charged black holes , '' proc .
a * 372 * ( 1980 ) 535 . g. w. gibbons , private communication . | a weyl solution describing two schwarzschild black holes is considered .
we focus on the @xmath0 invariant solution , with adm mass @xmath1 , where @xmath2 is the komar mass of each black hole . for this solution
the set of fixed points of the discrete symmetry is a totally geodesic sub - manifold .
the existence and radii of circular photon orbits in this sub - manifold are studied , as functions of the distance @xmath3 between the two black holes . for @xmath4
there are two such orbits , corresponding to @xmath5 and @xmath6 in schwarzschild coordinates . as the distance increases
, it is shown that the two photon orbits approach one another and merge when @xmath7 , where @xmath8 is the golden ratio . beyond this distance
there exist no circular photon orbits .
the two null orbits delimit a forbidden band for time - like circular orbits , which is interpreted in terms of optical geometry . for large @xmath9 ,
time - like circular orbits are allowed everywhere , as in the analogous newtonian problem .
the analysis is generalised by considering a @xmath0 invariant weyl solution with an array of @xmath10 black holes and also by charging the black holes , which connects the weyl solution to a majumdar - papapetrou spacetime . |
the goal of this paper is to derive an effective string theory of vortices beginning with a field theory containing classical vortex solutions .
the abelian higgs model is an example of such a theory .
nielsen and olesen @xcite showed that this model has classical magnetic vortex solutions .
these vortices are tubes of magnetic flux with constant energy per unit length .
the motivation for this work came from the dual superconductor picture of confinement @xcite . in this picture ,
a dual meissner effect confines electric color flux ( @xmath0 flux ) to narrow tubes connecting quark antiquark pairs .
calculations with explicit models of this type @xcite have been compared both with experimental data and with monte carlo simulations of qcd @xcite . to a good approximation , aside from a color factor , the dual abelian higgs model , coupling dual potentials to a scalar higgs field carrying magnetic charge , can be used to describe the results of these calculations . however , these calculations neglect the effect of fluctuations in the shape of the flux tube on the @xmath1 interactions . we show in this paper that taking account of those fluctuations leads to an effective string theory of long distance qcd .
well before the introduction of the idea of dual superconductivity , string models @xcite had been used to understand the origin of regge trajectories , and they have continued to be used to describe other features of hadron physics , such as the spectrum of hybrid mesons . in the dual superconductor picture , a string arises because the dual potentials couple to a quark
antiquark pair via a dirac string whose ends are a source and sink of electric color flux . the effect of the string is to create a flux tube ( or abrikosov nielsen olesen vortex @xcite ) connecting the quark
antiquark pair . as the pair moves
, this flux tube sweeps out a space time surface on which the dual higgs field must vanish .
this condition determines the location of the qcd string in the dual superconductor picture .
the effort to obtain an effective string theory for abrikosov nielsen olesen vortices has a long history , independent of any connection to qcd .
nambu @xcite attached quarks to the ends of superconducting vortices , and found an expression for the classical action of the resulting ano vortex in the singular london limit of infinite higgs mass .
he introduced a cutoff to render this action finite , and showed that it was proportional to the area of the worldsheet ( the nambu
goto action ) .
frster @xcite took into consideration the curvature of the worldsheet .
he showed that in the strong coupling limit , with the ratio of vector and scalar masses held fixed , the effects of curvature were unimportant , and the classical action for the vortex reduced to the nambu
goto action .
this limit can be regarded as the long distance limit , since only zero mass excitations are left in the theory .
equivalently , since the flux tube radius vanishes in this limit , all physical distances , measured in units of the flux tube radius , are becoming large .
all degrees of freedom except the transverse oscillations of the vortex are frozen out .
gervais and sakita @xcite first considered the quantum theory of the vortices of the abelian higgs model in the same long distance limit .
they used the results of frster to define collective coordinates for the vortices , by means of which they constructed an effective vortex action .
they also obtained a formal expression for the feynman path integral of the abelian higgs model as an integration over vortex sheets .
however , they were not able to write this expression as an integral over the physical degrees of freedom of the vortices .
lscher , symanzik , and weisz @xcite considered the leading semiclassical corrections to the classical nambu
goto action due to transverse string fluctuations , and showed how to regulate the resulting divergences .
they showed that for a string of length @xmath2 with fixed ends , the leading semiclassical contribution to the heavy quark potential is @xmath3 . in a second paper , lscher @xcite showed that this result was unaffected by the addition of other terms to the effective string action .
polchinski and strominger @xcite discussed the relation of the abelian higgs model to fundamental string theory , regarding the theory of ano vortices as an effective string theory .
they explained how existing string quantization methods were inappropriate for quantizing the vortices . to compensate for the anomalies @xcite in these quantization methods
, they introduced an additional term , the `` polchinski strominger term , '' into the effective vortex action .
akhmedov , chernodub , polikarpov , and zubkov @xcite studied the quantum theory of ano vortices in the london limit , in particular , they studied the transformation from field degrees of freedom to vortex degrees of freedom .
they showed that the jacobian of this transformation contained the `` polchinski strominger term '' as a factor .
although they , like gervais and sakita , did not obtain a complete expression for the path integral , this paper provided an important stimulus to our own work .
in the current paper , we simplify and extend work done in an earlier paper @xcite .
we begin with the path integral representation of a field theory having vortex solutions .
it is an effective field theory describing phenomena at distances greater than the flux tube radius .
we end up with an effective string theory of vortices in a form suitable for explicit calculations .
we apply this theory to calculate the energy @xmath4 and angular momentum @xmath5 of the fluctuations of a string bounded by the curve generated by the worldlines of a quark
antiquark pair separated by a fixed distance and rotating with fixed angular velocity .
this gives the contribution of string fluctuations to the regge trajectory @xmath6 , which we compare with the experimental @xmath7 and @xmath8 trajectories .
in section [ partition section ] , we rewrite the path integral over field configurations of the abelian higgs model containing vortices as an integral over surfaces on which the higgs field vanishes .
this introduces a jacobian due to the change from field variables to string variables ( surfaces ) .
this jacobian is the key to determining the action of the effective string theory , and to defining the integral over all surfaces .
we next use the formalism described in section [ partition section ] to obtain an effective theory of ano vortices . in section [ jacobian factor ] , we show how the jacobian divides into a field part and a string part . the two parts of the jacobian play different roles in the effective theory . in section
[ effective action section ] , we define an expression for the action of the effective string theory . all the dependence on the abelian higgs model is contained in the string action .
we also obtain an expression for the path integral over vortices . in section
[ coordinate fix ] , we show how to express the integral over surfaces as an integral over the two physical degrees of freedom of the vortex , and obtain the final form of the effective string theory . in the remaining sections we compute the leading semiclassical contribution to regge trajectories due to the fluctuations of the string .
we obtain an expression for the contribution of string fluctuations to the effective action in section [ small f section ] , and in sections [ regularization section ] and [ w_2 section ] describe how to regularize this expression , making use of the results of lscher , symanzik , and weisz @xcite . in section [ regge effective potential section ]
we calculate the contribution of string fluctuations to the effective action for a straight , rotating string , and in section [ regge plot section ] obtain the resulting corrections to regge trajectories .
in this section we consider the abelian higgs model coupled via a dirac string to a moving quark
antiquark pair .
we transform the path integral over field configurations containing vortices to an integral over the surfaces @xmath9 determining the location of the vortices .
we denote the ( dual ) potentials by @xmath10 and the complex ( monopole ) higgs field by @xmath11 .
the dual coupling constant is @xmath12 , where @xmath13 is the yang
mills coupling constant ( @xmath14 ) .
the worldlines of the quark and antiquark trajectories form the closed loop @xmath15 ( see fig .
[ loop figure ] ) .
= 1.8 in = 1.8 in the moving quark
antiquark pair couples to the dual potentials @xmath10 via a dirac string tensor @xmath16 , which is nonvanishing along some line @xmath17 connecting the @xmath1 pair
. as the pair moves , the line @xmath17 sweeps out a worldsheet @xmath18 parameterized by coordinates @xmath19 , @xmath20 .
the field @xmath11 vanishes on this worldsheet , @xmath21 the corresponding dirac string tensor @xmath16 is given by @xmath22 the action @xmath23 of a field configuration which has a vortex on the sheet @xmath18 is @xmath24 \ , , \label{firstaction}\ ] ] where the field strength @xmath25 is given by @xmath26 the higgs mechanism gives the vector particle ( dual gluon ) a mass @xmath27 and the scalar particle a mass @xmath28 , where @xmath29 is the vacuum expectation value of the higgs field .
we have introduced the color factor @xmath30 in ( [ g mu nu def ] ) because we are interested in using @xmath23 as a model for long distance qcd .
we consider @xmath23 to be an effective action describing distances greater than the flux tube radius @xmath31 .
the long distance @xmath1 interaction is determined by the wilson loop @xmath32 $ ] , @xmath33 = \int \scrd\phi^ * \scrd\phi \scrd c^{\mu } e^{i(s[\phi , c ] + s_{gf } ) } \ , , \label{originalpartition}\ ] ] where @xmath34 is a gauge fixing term
. the functional integrals are cut off at the momentum scale @xmath35 .
the action ( [ firstaction ] ) describes a field theory having classical vortex solutions .
the functional integral ( [ originalpartition ] ) goes over all field configurations containing a vortex bounded by @xmath15 .
previous calculations @xcite of @xmath32 $ ] were carried out in the classical approximation ( corresponding to a flat vortex sheet @xmath9 ) , and showed that the landau
ginzburg parameter @xmath36 is approximately equal to @xmath37 .
this corresponds to a superconductor on the border between type i and type ii .
in this situation , both particles have the same mass @xmath38 , the string tension is @xmath39 , and the flux tube radius is @xmath40 . to take into account the fluctuations of these vortices
, we must evaluate @xmath32 $ ] beyond the classical approximation .
we carry out the functional integration ( [ originalpartition ] ) in two steps : ( 1 ) we fix the location of a vortex sheet @xmath9 , and integrate only over field configurations for which @xmath41 vanishes on @xmath9 .
( 2 ) we integrate over all possible vortex sheets . to implement this procedure ,
we introduce into the functional integral ( [ originalpartition ] ) the factor one , written in the form @xmath42 \int \scrd \tilde x^\mu \delta\left[\re\phi(\tilde x^{\mu}(\xi))\right ] \delta\left[\im\phi(\tilde x^{\mu}(\xi))\right ] \ , .
\label{inserttildex}\ ] ] the integration @xmath43 is over the four functions @xmath18 .
the functions @xmath18 are a particular parameterization of the worldsheet @xmath9 .
the expression ( [ inserttildex ] ) implies that the string worldsheet @xmath9 , determined by the @xmath44 functions , is the surface of the zeros of the field @xmath11 .
the factor @xmath45 $ ] is a jacobian , and is defined by eq .
( [ inserttildex ] ) . inserting ( [ inserttildex ] ) into ( [ originalpartition ] ) puts the wilson loop in the form @xmath33 = \int \scrd\phi^ * \scrd\phi \scrd c^{\mu } e^{i(s[\phi , c ] + s_{gf } ) } j[\phi ] \int \scrd \tilde x^\mu \delta\left[\re\phi(\tilde x^{\mu}(\xi))\right ] \delta\left[\im\phi(\tilde x^{\mu}(\xi))\right ] \ , .
\label{beforeswitch}\ ] ] we then reverse the order of the field integration and the string integration over surfaces @xmath18
, @xmath33 = \int \scrd \tilde x^\mu \int \scrd\phi^ * \scrd\phi \scrd c^{\mu } j[\phi ] \delta\left[\re\phi(\tilde x^{\mu}(\xi))\right ] \delta\left[\im\phi(\tilde x^{\mu}(\xi))\right ] e^{i(s[\phi , c ] + s_{gf } ) } \ , . \label{afterswitch}\ ] ] in eq .
( [ beforeswitch ] ) , the @xmath44 functions fix @xmath9 to lie on the surface of the zeros of a given field @xmath11 , while in eq .
( [ afterswitch ] ) , they restrict the field @xmath11 to vanish on a given surface @xmath9 .
the integral over @xmath11 in eq .
( [ afterswitch ] ) is therefore restricted to functions @xmath11 which vanish on @xmath9 , in contrast to the integral over @xmath11 in eq .
( [ beforeswitch ] ) , in which @xmath11 can be any function .
to evaluate @xmath32 $ ] we divide @xmath45 $ ] into two parts . the jacobian @xmath45 $ ] in eq .
( [ afterswitch ] ) is evaluated for field configurations @xmath11 which vanish on a particular surface @xmath9 .
we make this explicit by writing ( [ inserttildex ] ) as @xmath46^{-1 } = \int \scrd \tilde y^\mu \delta\left[\re\phi(\tilde y^{\mu}(\tau))\right ] \delta\left[\im\phi(\tilde y^{\mu}(\tau))\right ] \ , , \label{jdef}\ ] ] where @xmath47 is some other string worldsheet , distinct from @xmath9 .
the evaluation of the jacobian is the essential new ingredient in deriving @xmath32 $ ] . the @xmath44 functions in ( [ jdef ] ) select surfaces @xmath48 which lie in a neighborhood of the surface @xmath18 of the zeros of @xmath11 .
we separate @xmath48 into components lying on the surface @xmath18 and components lying along vectors @xmath49 normal to @xmath18 at the point @xmath50 : @xmath51 the point @xmath52 is the point on the surface @xmath18 lying closest to @xmath48 , and the magnitude of @xmath53 is the distance from @xmath48 to @xmath52 ( see fig .
[ normal figure ] ) .
we evaluate the jacobian ( [ jdef ] ) by making the change of variables @xmath54 defined by ( [ localcoords ] ) . although the @xmath44 functions in ( [ jdef ] ) force @xmath55 to vanish , the integrations over @xmath55 give a contribution to the jacobian .
furthermore , this contribution depends on the field variable @xmath11 in a neighborhood of the surface .
the integration over the reparameterizations @xmath56 of the surface @xmath18 , on the other hand , depends upon the surface , but not on the fields .
the change of variables ( [ transformation ] ) leads to a factorization of the jacobian into a field contribution , and into a contribution depending only on the intrinsic properties of the worldsheet @xmath57 .
we now exhibit the factorization of the jacobian . under the transformation ( [ transformation ] ) , the integral over @xmath47 becomes @xmath58 \scrd
y_\perp^a \scrd\xi \nonumber \\ & = & \det_\tau\left[\sqrt { -\frac{1}{2 } \left ( \epsilon^{ab } \frac{\partial \tilde x^\mu}{\partial\xi^a } \frac{\partial \tilde x^\nu}{\partial\xi^b } \right)^2 \frac{1}{2 } \left ( \epsilon^{ab } n_\alpha^a n_\beta^b \right)^2}\right ] \scrd y_\perp^a \scrd\xi \nonumber \\ &
= & \det_\tau\left[\sqrt{-g(\xi)}\big|_{\xi=\xi(\tau)}\right ] \scrd y_\perp^a \scrd\xi \ , , \label{y jacobian}\end{aligned}\ ] ] where @xmath59 is the square root of the determinant of the induced metric @xmath60 evaluated on the worldsheet @xmath9 .
appendix [ notation appendix ] gives a summary of our notation , and of the relations used to obtain ( [ y jacobian ] ) .
the functional determinant in ( [ y jacobian ] ) is the product of its argument evaluated at all points @xmath61 on the sheet , in the same way that the integration over @xmath62 is a product of integrals at all points @xmath61 . making the change of coordinates
( [ localcoords ] ) , ( [ transformation ] ) in the jacobian ( [ jdef ] ) gives @xmath63^{-1 } & = & \int \scrd \xi \scrd y_{\perp}^a \det_\tau[\sqrt{-g } ] \delta\left[\re\phi\left(\tilde x^{\mu}(\xi(\tau ) ) + y_{\perp}^a(\xi(\tau ) ) n_a^{\mu}(\xi(\tau))\right)\right ] \nonumber \\ & & \times \delta\left[\im\phi\left(\tilde x^{\mu}(\xi(\tau ) ) + y_{\perp}^a(\xi(\tau ) ) n_a^{\mu}(\xi(\tau))\right)\right ] \ , .
\label{j partial split}\end{aligned}\ ] ] eq .
( [ j partial split ] ) has the form : @xmath64^{-1 } = \int \scrd \xi(\tau ) \det_{\tau}\left[\sqrt{-g}\right ] j_{\perp}[\phi , \tilde x^{\mu}(\xi(\tau))]^{-1 } \
, , \label{newjacob}\ ] ] where @xmath65^{-1 } & = & \int \scrd y_{\perp}^a \delta\left[\re\phi\left(\tilde x^{\mu}(\xi(\tau ) ) + y_{\perp}^a(\xi(\tau ) ) n_a^{\mu}(\xi(\tau))\right)\right ] \nonumber \\ & & \times \delta\left[\im\phi\left(\tilde x^{\mu}(\xi ( \tau ) ) + y_{\perp}^a(\xi(\tau ) ) n_a^{\mu}(\xi(\tau))\right)\right]\end{aligned}\ ] ] contains all the dependence on @xmath11 . since @xmath66 is independent of the parameterization @xmath56 , the jacobian factors into two parts : @xmath64^{-1 } = j_{\parallel}[\tilde x]^{-1 } j_{\perp}[\phi , \tilde x]^{-1 } \ , , \label{jfactor}\ ] ] where @xmath67^{-1 } = \int \scrd \xi \det_{\tau}\left[\sqrt{-g}\right ] \ , .
\label{jparallel}\ ] ] the string part @xmath68 of the jacobian arises from the parameterization degrees of freedom . in the next section ,
we show that @xmath66 is the faddeev
popov determinant for the @xmath44 functions in ( [ afterswitch ] ) .
this allows us to define the action of the effective string theory . in the following section
, we will use @xmath68 to fix the reparameterization degrees of freedom .
inserting the factorized form ( [ jfactor ] ) of @xmath45 $ ] into the expression ( [ afterswitch ] ) for @xmath32 $ ] gives the wilson loop the form @xmath33 = \int \scrd \tilde x^\mu j_\parallel[\tilde x ] e^{is_{{\rm eff } } } \ , , \label{stringrep}\ ] ] where the action @xmath69 of the effective string theory is given by @xmath70 } = \int \scrd\phi^ * \scrd\phi \scrd c^\mu j_\perp[\phi ] \delta\left[\re\phi(\tilde x^\mu(\xi))\right ] \delta\left[\im\phi(\tilde x^\mu(\xi))\right ] e^{i(s + s_{gf } ) } \ , .
\label{seff}\ ] ] the string action ( [ seff ] ) was obtained previously by gervais and sakita @xcite .
the novel feature of our result is the string integration measure of the wilson loop ( [ stringrep ] ) .
the string action depends upon the field part @xmath66 of the jacobian , @xmath71^{-1 } = \int \scrd y_{\perp}^a \delta\left[\re\phi\left(\tilde x^{\mu } + y_{\perp}^a n_{\mu a } \right)\right ] \delta\left[\im\phi\left(\tilde x^{\mu } + y_{\perp}^a n_{\mu a } \right)\right ] \ , .
\label{j perp of y}\ ] ] the @xmath44 functions force @xmath72 to be zero , so we can expand their arguments in a power series in @xmath72 , @xmath73 the zeroth order term in ( [ phi expand in y ] ) vanishes because @xmath74 is the surface of the zeros of @xmath11 . the integration ( [ j perp of y ] ) over @xmath55
gives the result @xmath71^{-1 } = \det^{-1}_{\xi}\left [ \epsilon^{ab } n_a^{\mu } n_b^{\nu } \left(\partial_{\mu}\re\phi\right ) \left(\partial_{\nu}\im\phi\right ) \big|_{x^{\mu } = \tilde x^{\mu } } \right ] \ , .
\label{j_perp expressed}\ ] ] the jacobian @xmath66 is a faddeev
popov determinant , which we discuss in appendix [ j_perp fp appendix ] .
( [ seff ] ) gives the action @xmath75 of the effective string theory as an integral over field configurations which have a vortex fixed at @xmath9 .
since the vortex theory ( [ originalpartition ] ) is an effective long distance theory , the path integral ( [ originalpartition ] ) for @xmath32 $ ] , written in terms of the fields of the abelian higgs model , is cut off at a scale @xmath76 which is on the order of the mass @xmath77 of the dual gluon .
furthermore , the integration ( [ stringrep ] ) over @xmath9 includes all the long distance fluctuations of the theory .
therefore , the path integral ( [ seff ] ) contains neither short distance nor long distance fluctuations , and is determined by minimizing the field action @xmath78 $ ] for a fixed position of the vortex sheet : @xmath79 = s[\tilde x^\mu , \phi^{{\rm class } } , c_\mu^{{\rm class } } ] \ , , \kern 0.5 in \phi^{{\rm class}}(\tilde x^\mu ) = 0 \ , .
\label{classical string}\ ] ] the fields @xmath80 and @xmath81 are the solutions of the classical equations of motion , subject to the boundary condition @xmath82 . the action @xmath69 depends both on the distance @xmath2 between the quarks , and the radius of curvature @xmath83 of the vortex sheet bounded by @xmath15 .
in the long distance limit , when the length of the string @xmath2 and its radius @xmath83 are large compared to the thickness of the flux tube @xmath31 , the string action ( [ classical string ] ) becomes the nambu
goto action @xmath84 , @xmath85 where @xmath86 is the classical string tension , determined from the solution of the nielsen
olesen equations for a straight , infinitely long string .
it is convenient to separate the action ( [ classical string ] ) into its perturbative and nonperturbative parts : @xmath79 = s[\tilde x^\mu , \phi^{{\rm class } } , c_\mu^{{\rm class } } ] = s^{{\rm maxwell}}[\tilde x^\mu ] + s^{{\rm np}}[\tilde x^\mu ] \
, , \label{break abelian action}\ ] ] where @xmath87 is the action obtained by setting @xmath88 in eq .
( [ firstaction ] ) .
the value of @xmath87 depends only upon the boundary @xmath15 , and is the usual electromagnetic interaction between charged particles : @xmath89 = \frac{4}{3 } \frac{e^2}{2 } \oint dx^\mu \oint d{x'}^\mu { \cal d}_{\mu\nu}(x^\mu - { x'}^\mu ) \
, , \label{maxwell action}\ ] ] where @xmath90 is the photon propagator . to calculate the wilson loop @xmath32 $ ] from the effective string theory ( [ stringrep ] )
, we must also examine @xmath69 at smaller values of @xmath2 and @xmath83 , on the order of the string thickness @xmath31 .
we first consider the dependence of @xmath69 on @xmath2 for a flat string , where @xmath91 . in this case , the curve @xmath15 is a rectangle of length @xmath92 in the time direction , and width @xmath2 in the space direction . in the large @xmath92 limit , the action @xmath69 reduces to the product of @xmath92 and the potential @xmath93 previously used to fit the energy levels of heavy quark systems .
evaluation of ( [ break abelian action ] ) for a flat sheet gives a corresponding decomposition of @xmath93 , @xmath94 for small @xmath2 , @xmath95 while eq .
( [ ng action def ] ) gives the large @xmath2 behavior @xmath96 recent numerical studies @xcite of the classical equations of motion for a flat sheet have shown that for a superconductor on the i ii border , the long distance behavior ( [ v class long ] ) of @xmath97 persists to small values of @xmath2 , even to values less than the string thickness @xmath31 .
therefore , for a superconductor on the i ii border , @xmath93 is , to a good approximation , equal to the cornell potential @xcite : @xmath98 in other words , for a flat sheet , @xmath99 thus , for short straight strings the nambu
goto action remains a good approximation to the nonperturbative part of the classical action for a superconductor on the type i ii border .
next , consider the nonperturbative contribution to the classical action for a long bent string .
( the maxwell action has the value ( [ maxwell action ] ) independent of the shape of the vortex . ) the leading correction to the nambu
goto action when the string is bent is the curvature term : @xmath100 where @xmath101 is the extrinsic curvature .
@xmath102 the magnitude of @xmath101 is of the order of @xmath103 , so that @xmath104 .
the calculation of the `` rigidity '' @xmath105 determining the size of @xmath106 has been considered by a number of authors @xcite , but the value of @xmath105 for a superconductor on the i ii border was never calculated .
we conjecture that the value of @xmath105 is small , because de vega and schaposnik @xcite have shown that the components of the stress tensor perpendicular to the axis of a straight nielsen olesen flux tube vanish for a superconductor on the border between type i and type ii . in other words , there are no `` bonds '' perpendicular to the field lines of a straight flux tube of infinite extent .
when the flux tube is bent slightly , there are no perpendicular bonds to be stretched or compressed , and the change in the energy is just the string tension multiplied by the change in length .
that is , the curvature term , which in a sense represents the attraction or repulsion between neighboring parts of the string , should vanish .
a more formal argument can be made by regarding the borderline superconductor as the long distance limit of a theory where the forces between vortices become weak .
polyakov @xcite has shown , using renormalization group methods , that @xmath105 also vanishes in this limit .
similar heuristic arguments give a reason for the above mentioned result that the nambu
goto action is a good approximation for short , straight strings on the i ii border .
the bending of the field lines as the quark antiquark separation becomes smaller causes no additional changes in the energy .
we therefore take the action of the effective string theory to be equal to the sum of the maxwell action ( [ maxwell action ] ) and the nambu
goto action ( [ ng action def ] ) : @xmath79 = s^{{\rm maxwell}}[\gamma ] - \sigma \int d^2\xi \sqrt{-g } \ , .
\label{given action}\ ] ] we use eq .
( [ given action ] ) for the full range of string lengths @xmath2 and radii of curvature @xmath83 greater than the inverse of the mass @xmath77 of the dual gluon , which is the cutoff of the effective string theory ( [ stringrep ] ) .
( [ given action ] ) for @xmath107 $ ] is the generalization of ( [ v class ] ) to a general sheet . the first term , @xmath108 $ ] , is just a boundary term , independent of the fluctuating string , and we take take @xmath109 for the calculations carried out in the rest of this paper .
in the next section we show how to carry out the integration over @xmath9 in ( [ stringrep ] ) by separating the degrees of freedom of the worldsheet @xmath9 into two physical degrees of freedom and two reparameterization degrees of freedom .
this treatment makes no use of ( [ given action ] ) , and is applicable to any effective string theory of vortices .
we next show how to evaluate the integral over @xmath18 in eq .
( [ stringrep ] ) , @xmath33 = \int \scrd \tilde x^\mu j_\parallel[\tilde x^\mu ] e^{is_{{\rm eff } } } \ , .
\label{vortex partition}\ ] ] the integration over @xmath18 is the product of an integral over string worldsheets and an integral over reparameterizations of the coordinates of the string .
the jacobian @xmath68 is the inverse of the integration ( [ jparallel ] ) over reparameterization degrees of freedom . in this section ,
we fix the parameterization of the string , and show that @xmath68 cancels the integration over reparameterizations .
any surface @xmath9 has only two physical degrees of freedom .
the other two degrees of freedom represent the invariance of the surface under coordinate reparameterizations .
we fix the coordinate reparameterization symmetry by choosing a particular `` representation '' @xmath110 of the surface , which depends on two functions @xmath111 , @xmath112 , @xmath113 \ , .
\label{x_p def}\ ] ] a particular example of a representation @xmath110 is obtained by expanding in transverse fluctuations @xmath114 about a fixed sheet @xmath115 , @xmath116 = \bar x_p^\mu(\xi ) + x_\perp^a(\xi ) \bar n^\mu_a(\xi ) \ , .
\label{normal rep}\ ] ] the vectors @xmath117 are orthogonal to the surface @xmath115 . in this example , @xmath118 and @xmath119 are the transverse coordinates @xmath114
. any physical surface can be expressed in terms of @xmath110 by a suitable choice of @xmath118 and @xmath119 . in particular , the worldsheet @xmath18 appearing in the integral ( [ vortex partition ] ) can be written in terms of a reparameterization @xmath120 of the representation @xmath110 , @xmath121 \ , .
\label{x tilde of x_p}\ ] ] the four degrees of freedom in @xmath18 are replaced by two physical degrees of freedom @xmath111 , @xmath112 and two reparameterization degrees of freedom @xmath120 .
we can write the integral over @xmath18 in ( [ vortex partition ] ) in terms of integrals over @xmath122 and @xmath120 , @xmath123 \scrd f^1 \scrd f^2 \scrd\tilde\xi \,.\ ] ] noting that the derivative of @xmath9 with respect to @xmath120 is , @xmath124 \bigg|_{\xi=\tilde\xi } \,,\ ] ] we can write @xmath125 \scrd f^1 \scrd f^2 \scrd \tilde\xi \nonumber \\ & = & \det\left [ \tilde t_{\mu\nu } \sqrt{-g^p } \frac{\partial x_p^\mu}{\partial f^1 } \frac{\partial x_p^\nu } { \partial f^2 } \bigg|_{\xi = \tilde\xi } \right ] \scrd f^1 \scrd f^2 \scrd \tilde\xi \ , , \label{measure shift}\end{aligned}\ ] ] where @xmath126 is the antisymmetric tensor normal to the worldsheet and , @xmath127 is the induced metric of @xmath128 .
the metric @xmath129 is related to the metric @xmath130 of @xmath18 , @xmath131 the induced metric @xmath130 of the original worldsheet @xmath18 does not appear in eq .
( [ measure shift ] ) because the determinant is independent of @xmath132 . only the induced metric @xmath129 of the worldsheet @xmath128 enters into the determinant . with the parameterization ( [ x tilde of x_p ] ) of @xmath9
, the path integral ( [ vortex partition ] ) takes the form @xmath33 = \int \scrd\tilde\xi \scrd f^1 \scrd f^2 \det\left [ \tilde t^{\mu\nu } \frac{\partial x_p^\mu } { \partial f^1 } \frac{\partial x_p^\nu } { \partial f^2 } \right ] \det[\sqrt{-g^p } ] j_\parallel e^{is_{{\rm eff } } } \ , .
\label{w subbed}\ ] ] the action @xmath69 is parameterization independent , so it is independent of @xmath120 .
the same is true for @xmath68 .
furthermore , @xmath133 is parameterization independent , so that the product @xmath134\ ] ] is independent of @xmath120 .
therefore , this product , along with @xmath68 and @xmath135 , can be brought outside the @xmath132 integral in ( [ w subbed ] ) .
the path integral then takes the form @xmath33 = \int \scrd f^1 \scrd f^2 \det\left [ \tilde t^{\mu\nu } \frac{\partial x_p^\mu } { \partial f^1 } \frac{\partial x_p^\nu } { \partial f^2 } \right ] j_\parallel e^{is_{{\rm eff } } } \int \scrd \tilde\xi \det[\sqrt{-g^p } ] \ , .
\label{before cancel j parallel}\ ] ] the remaining integral over reparameterizations @xmath132 is equal to @xmath136 , defined by ( [ jparallel ] ) , and is canceled by the explicit factor of @xmath68 appearing in ( [ before cancel j parallel ] ) .
this means we do not need to evaluate @xmath68 , and can avoid the complications inherent in evaluating the integral over reparameterizations of the string coordinates .
the anomalies produced in string theory by evaluating this integral are not present , so we do not have a polchinski strominger term in the theory .
( [ before cancel j parallel ] ) gives the final result for the wilson loop @xmath33 = \int \scrd f^1 \scrd f^2 \det\left [ \tilde t_{\mu\nu } \frac{\partial x_p^\mu}{\partial f^1 } \frac{\partial x_p^\nu } { \partial f^2 } \right ] e^{is_{{\rm eff } } } \ , , \label{param measure}\ ] ] as an integration over two function @xmath111 and @xmath112 , the physical degrees of freedom of the string .
the path integral ( [ param measure ] ) is invariant under reparameterizations of the string , and describes a two dimensional field theory with two degrees of freedom , the two transverse oscillations of a two dimensional sheet .
the integration ( [ param measure ] ) goes over the normal fluctuations of the string worldsheet .
the components of @xmath118 and @xmath119 along the sheet are nonphysical .
the determinant in ( [ param measure ] ) is a normalization factor for @xmath118 and @xmath119 .
this can be seen by applying the identity @xmath137 to the determinant , @xmath138 = \det\left [ \epsilon^{ab } \left ( n_{\mu a } \frac{\partial x_p^\mu } { \partial f^1 } \right ) \left ( n_{\nu b } \frac{\partial x_p^\nu } { \partial f^2 } \right ) \right ] \,.\ ] ] the factors of @xmath139 determine the amount of the fluctuation @xmath140 which is in a direction normal to the sheet .
( [ param measure ] ) is the string representation of any field theory containing classical vortex solutions .
the expression ( [ seff ] ) for @xmath69 had been obtained previously by gervais and sakita @xcite ; we are unaware of any previous derivation of the string representation ( [ param measure ] ) for the path integral .
we now show how it provides a method for explicit calculations .
in this section we carry out the semiclassical expansion of @xmath32 $ ] about a classical solution of the effective string theory , and find the leading contribution of string fluctuations to the effective action @xmath141 $ ] .
the ends of the string follow the path @xmath15 fixed by the prescribed trajectory of the quarks , and the fluctuations of the string are cutoff at the momentum scale @xmath77 of the inverse string radius .
as explained in section [ effective action section ] , we take the action @xmath69 of the effective string theory to be the nambu goto action @xmath142 and ( [ param measure ] ) becomes @xmath33 = \int \scrd f^1(\xi ) \scrd f^2(\xi ) \det\left [ \tilde t^{\mu\nu } \frac{\partial x_p^\mu}{\partial f^1 } \frac{\partial x_p^\nu}{\partial f^2 } \right ] e^{-i\sigma \int d^2\xi \sqrt{-g } } \ , .
\label{w with action}\ ] ] we expand ( [ w with action ] ) in small fluctuations @xmath140 of @xmath143 $ ] around a fixed sheet @xmath144 , subject to the condition that the boundary of @xmath115 lies on the curve @xmath15 , @xmath145 where @xmath146 is the position of the string worldsheet when @xmath147 .
expanding @xmath59 to quadratic order in small @xmath118 and @xmath119 , we obtain @xmath148 & = & \int \scrd f^i(\xi ) \det\left [ \frac{1}{2 } \tilde t^{\mu\nu } \epsilon^{ij } \frac{\partial x_p^\mu}{\partial f^i } \frac{\partial x_p^\nu}{\partial f^j } \bigg|_{f^i=0 } \right ] \nonumber \\ & & \times \exp\left\ { -i\sigma \int d^2\xi \sqrt{-\bar g } \left [ 1 + \bar g^{ab } \frac{\partial x_p^\mu}{\partial\xi^a } \bigg|_{f^i=0 } \frac{\partial}{\partial\xi^b } \left ( \frac{\partial x_p^\mu } { \partial f^i } \bigg|_{f^i=0 } f^i \right ) + \frac{1}{2 } f^i g^{-1}_{ij } f^j \right ] \right\ } \ , , \label{semiclassical string partition}\end{aligned}\ ] ] where @xmath149 is @xmath150 the metric of the fixed worldsheet @xmath115 , and @xmath151 we choose @xmath115 to be the surface which minimizes the action .
then @xmath115 satisfies the `` classical equation of motion '' @xmath152 where the covariant laplacian is @xmath153 using the fact that the covariant derivative of the metric is zero , we show in appendix [ notation appendix ] that @xmath154 the vectors @xmath155 and @xmath156 form a complete basis , so ( [ classical string equation ] ) and ( [ covariant identity ] ) imply @xmath157 evaluating the @xmath140 integral in eq .
( [ semiclassical string partition ] ) gives @xmath33 = e^{-i\sigma \int d^2\xi \sqrt{-\bar g } } \det\left [ \frac{1}{2 } \tilde t_{\mu\nu } \epsilon^{ij } \frac{\partial x_p^\mu}{\partial f^i } \frac{\partial x_p^\nu}{\partial f^j } \bigg|_{f^i=0 } \right ] \det^{-1/2}\left[g^{-1}_{ij}\right ] \ , .
\label{z of g}\ ] ] the inverse propagator @xmath158 ( [ g^-1_ij def ] ) can be shown to be @xmath159 \bar n_{\nu b } \frac{\partial x_p^\nu}{\partial f^j } \bigg|_{f^i=0 } \ , , \label{g inverse def}\ ] ] where @xmath160 is the extrinsic curvature tensor of the sheet @xmath115
. a derivation of ( [ g inverse def ] ) is given in appendix a of @xcite .
the @xmath161 are vectors normal to the worldsheet @xmath115 .
( [ g inverse def ] ) gives @xmath162 = \det^{-1/2}\left [ -\nabla^2 \delta_{ab } - \bar\scrk^a_{ab } \bar\scrk^{bab } \right ] \det^{-1}\left[\frac{1}{2 } \epsilon^{ab } \bar n_{\mu a } \bar n_{\nu b } \epsilon^{ij } \frac{\partial x_p^\mu}{\partial f^i } \frac{\partial x_p^\nu}{\partial f^j } \bigg|_{f^i=0 } \right ] \ , .
\label{g det}\ ] ] from the identity ( [ dual normal ] ) , @xmath163 we see that the first determinant in ( [ z of g ] ) and the second determinant in ( [ g det ] ) cancel .
the determinant appearing in ( [ param measure ] ) produces exactly the correct normalization for the green s function .
the functional integral ( [ z of g ] ) becomes @xmath33 = e^{-i\sigma \int d^2 \xi \sqrt{-\bar g } } \det^{-1/2}\left [ -\nabla^2 \delta_{ab } - \bar\scrk^a_{ab } \bar\scrk^{bab } \right ] \ , .
\label{last z}\ ] ] we note that ( [ last z ] ) is independent of the factors of @xmath164 which appeared in the inverse propagator ( [ g inverse def ] ) .
these factors are the projections of the fluctuations @xmath140 normal to the string worldsheet . for small @xmath140 ,
the worldsheet @xmath110 is @xmath165 the perturbation of the worldsheet in the direction @xmath166 is @xmath167 the factors of @xmath168 in ( [ g inverse def ] ) project out the part of the fluctuations @xmath140 perpendicular to the worldsheet @xmath115 .
only normal fluctuations contribute to @xmath32 $ ] , since fluctuations along the worldsheet are equivalent to a reparameterization of the sheet coordinates . the effective action obtained from ( [ last z ] ) is @xmath169 = s_{cl } + s_{{\rm fluc } } \ , .
\label{effective action def}\ ] ] the first term in ( [ effective action def ] ) is the nambu
goto action evaluated at the `` classical '' worldsheet @xmath144 , @xmath170 the semiclassical correction @xmath171 due to the transverse string fluctuations is @xmath172 \ , .
\label{w unreg}\ ] ] to summarize , we have integrated out the string fluctuations , and reduced the problem to the evaluation of the determinant in ( [ last z ] ) .
this is a quantum mechanical scattering problem in the background of the solution of the classical equation ( [ class string simple ] ) , with appropriate boundary conditions . in the next section ,
we describe how to evaluate this determinant .
the argument of the logarithm in ( [ w unreg ] ) is the inverse propagator for fluctuations on the string .
this inverse propagator can also be obtained by direct variation of the nambu
goto action with respect to any transverse coordinates @xmath114 , @xmath173 up to an overall normalization factor .
in fact , the correction ( [ w unreg ] ) to the effective action has already been studied by lscher , symanzik , and weisz ( lsw ) @xcite in the case of a straight string with fixed ends .
we describe their results , which we will use in evaluating ( [ w unreg ] ) .
lsw used pauli
villars regularization to obtain a regulated form @xmath174 of the trace in ( [ w unreg ] ) , @xmath175 the @xmath176 are the masses of the regulators , and the @xmath177 are suitably chosen coefficients .
the laplacian in ( [ regulated action ] ) has been wick rotated from minkowski to euclidean space . the regulated quantity @xmath174 is separated into a divergent part @xmath178 and a finite part @xmath179 , @xmath180 lsw evaluated the divergent part @xmath181 , and obtained terms which are quadratically , linearly , and logarithmically divergent in the cutoffs @xmath176 .
the quadratic term is a renormalization of the string tension , the linear term is a renormalization of the quark masses , and the logarithmically divergent term is proportional to the integral over all space of the scalar curvature @xmath182 of the string worldsheet . lsw also obtained a formal expression for the finite part @xmath179 .
they evaluated this expression only for the case of a straight string of length @xmath2 with fixed ends , and calculated a correction @xmath183 to the static potential : @xmath184 we are interested in calculating @xmath171 for rotating quarks , so we must evaluate @xmath174 for a more general surface .
we break ( [ w unreg ] ) into two parts : @xmath185 + \frac{i}{2 } \tr\ln\left [ \frac{-\nabla^2 \delta_{ab } - \bar \scrk^a_{ab } \bar \scrk^{bab}}{-\nabla^2 } \right ] \ , .
\label{separated effective action}\ ] ] we will evaluate the first term in ( [ separated effective action ] ) by generalizing the calculation of lsw .
we will calculate the second term directly . the first term in ( [ separated effective action ] )
, @xmath186 \ , , \label{w_1}\ ] ] involves the laplacian in the curved background of the classical solution @xmath115 . in the flat case studied by lsw ,
the laplacian is equal to @xmath187 , @xmath188 the coordinate @xmath189 is the time in the lab frame , and @xmath190 is a radial coordinate which takes the values @xmath191 and @xmath192 at the two ends of the string .
the length of the string is @xmath193 . to calculate @xmath194 we extend the calculation of lsw to more general coordinate systems .
we make a coordinate transformation @xmath195 to conformal coordinates , where the transformed metric @xmath196 , @xmath197 .
this transformation puts the laplacian in a form similar to the flat sheet laplacian ( [ flat sheet laplacian ] ) , and allows us to evaluate ( [ w_1 ] ) by extending the calculation of lsw . to see how this works , we express @xmath198 as a functional integral , @xmath199 the transformation to conformal coordinates @xmath200 gives @xmath201 which is of the form of @xmath174 treated by lsw .
we will need @xmath194 in the limit of large @xmath92 , and hence are only interested in strings whose metric is time independent . to determine which metrics are time independent
, we must choose a coordinate system .
we choose coordinates @xmath190 and @xmath189 , where @xmath189 is the time in the lab frame , and @xmath190 is orthogonal to @xmath189 ( @xmath202 ) .
this guarantees that @xmath189 is the physical time . from now on we consider only metrics @xmath130 which are independent of @xmath189 . in appendix
[ luscher appendix ] , we show that @xmath203 where @xmath204 contains quark mass and string tension renormalizations .
the finite part of @xmath194 is @xmath205 where @xmath206 the results ( [ effective r proper ] ) and ( [ r_p def text ] ) are valid for any orthogonal coordinate system with a time independent metric .
we show in appendix [ luscher appendix ] that @xmath207 is equal to the classical energy of the string divided by the string tension @xmath86 .
we call @xmath207 the `` proper length '' of the string . for a flat metric , where @xmath208 , the proper length @xmath207 of the string reduces to the distance @xmath2 between its endpoints .
in the previous section , we evaluated the finite part of @xmath194 for a sheet with a time independent metric using the results of lsw . in this section
we evaluate the second term in ( [ separated effective action ] ) , @xmath209 \ , .
\label{w_2 def}\ ] ] the trace in ( [ w_2 def ] ) is over functions of @xmath190 and @xmath189 .
we first make the coordinate transformation @xmath210 , @xmath211 the coordinate @xmath212 runs from @xmath213 to @xmath214 , @xmath215 and @xmath216 . in appendix
[ luscher appendix ] , we show that the metric in the system @xmath217 is conformal ( @xmath218 , @xmath219 ) . in this coordinate system ,
the inverse propagator for string fluctuations is @xmath220 the string has infinite extent in time , and the curvature @xmath221 is independent of @xmath189 , so we can take the fourier transform with respect to the time coordinate .
we express the trace in ( [ w_2 def ] ) over functions of @xmath189 and @xmath190 as an integral over a frequency @xmath222 and a trace over functions of a single variable @xmath212 , @xmath223 \ , .
\label{w_2 of nu}\ ] ] in going from ( [ w_2 def ] ) to ( [ w_2 of nu ] ) , we have also carried out the wick rotation @xmath224 .
the integration over @xmath222 gives @xmath225 \ , .
\label{w_2 given}\ ] ] eq .
( [ w_2 given ] ) expresses @xmath226 as the trace of the difference of two operators .
the first has the form of a hamiltonian for a relativistic particle in the local potential @xmath227 .
the second operator has the form of a free hamiltonian .
the square roots enter because we are working with relativistic degrees of freedom .
the terms @xmath194 and @xmath226 are proportional to the time @xmath92 , but are otherwise time independent .
we define the `` effective lagrangian '' of the string to be the effective action divided by the time @xmath92 .
the sum of ( [ effective r proper ] ) and ( [ w_2 given ] ) gives the effective lagrangian @xmath228 determining the contribution of the string fluctuations to @xmath32 $ ] , @xmath229 in the next section , we will evaluate @xmath228 for a string of length @xmath2 rotating with angular velocity @xmath8 . in appendix
[ log cutoff appendix ] , we show that , for a general sheet , @xmath226 is logarithmically divergent .
we show that its divergent part is given by @xmath230 where @xmath182 is the scalar curvature , @xmath231 this result agrees , in the large time limit , with the logarithmically divergent term in the cutoff dependent part of the effective string action ( [ luscher reg ] ) found by lsw .
we now evaluate the effective lagrangian of a string with boundary @xmath15 generated by a quark
antiquark pair separated at fixed distances @xmath232 and @xmath192 from the origin , and rotating with angular velocity @xmath8 .
this lagrangian has two parts , the classical string lagrangian and the contribution ( [ rotating potential ] ) of string fluctuations .
we evaluate the classical string lagrangian first .
the solution to the classical equations of motion ( [ class string simple ] ) yields the classical , straight rotating string , @xmath233 the coordinate @xmath190 is chosen so that the velocity of the string is zero when @xmath234 .
the coordinate @xmath190 runs from @xmath191 to @xmath192 , and @xmath189 runs from @xmath235 to @xmath236 .
the vectors @xmath237 and @xmath238 are two orthogonal unit vectors in the plane of rotation , and @xmath239 is a unit vector in the time direction .
the classical lagrangian @xmath240 obtained from ( [ s_ng def ] ) is @xmath241 next , we calculate the contribution @xmath228 ( [ rotating potential ] ) due to string fluctuations .
the metric of the sheet ( [ x bar def ] ) is @xmath242 this metric is independent of @xmath189 , and @xmath243 .
we make the transformation ( [ coord x def ] ) from coordinates @xmath190 and @xmath189 to coordinates @xmath212 and @xmath189 , and find @xmath244 the coordinate @xmath212 runs from @xmath213 to @xmath214 , where @xmath245 and the proper length @xmath207 of the string is @xmath246 using this coordinate system , we evaluate @xmath228 in appendix [ eigenfunctions and sum ] for the case of equal quark masses , @xmath247 : @xmath248 - \frac{v^2 \gamma}{\pi r } f\left(v\right ) \ , , \label{v string}\ ] ] where @xmath249 and the function @xmath250 is @xmath251 \,.\ ] ] eq .
( [ v string ] ) becomes the lscher term in the zero velocity limit .
we are interested in the large @xmath2 limit , where the quark velocity is close to the speed of light . for @xmath252 close to one ,
( [ v string ] ) becomes @xmath253 + \frac{7}{6r } + o\left(\frac{\ln\gamma}{\gamma r}\right ) \ , .
\label{l fluc calculated}\ ] ] furthermore , for the semiclassical expansion to be valid , the theory must be weakly coupled .
that is , @xmath228 must be less than @xmath240 ( [ l classical string ] ) . for large @xmath2 ,
@xmath254 the semiclassical expansion is valid , since , as we will see , @xmath2 grows like @xmath255 in the @xmath256 limit . in this case , the long distance limit where the effective theory is applicable is automatically the region of weak coupling .
we calculate classical regge trajectories for equal mass quarks by adding a quark mass term to the string lagrangian @xmath240 , @xmath257 we have used eq .
( [ l classical string ] ) with @xmath247 .
the quark velocity is @xmath258 , and @xmath259 is the quark boost factor .
the lagrangian ( [ l_cl def ] ) is a function of @xmath2 and @xmath8 , @xmath260 the angular momentum of the meson is obtained by varying the lagrangian with respect to the angular velocity , @xmath261 the meson energy is given by the hamiltonian , @xmath262 the classical equation of motion @xmath263 for the quarks determines @xmath2 as a function of @xmath8 , @xmath264 eq . ( [ boundary condition ] ) shows that @xmath2 is proportional to @xmath255 for large @xmath265 . expanding ( [ j equation ] ) and
( [ e equation ] ) in the large @xmath2 limit , where the quark velocity @xmath252 goes to one , yields the result : @xmath266 the first term in ( [ classical regge eqn ] ) is the classical formula for the slope of a regge trajectory .
the second term is the leading correction for nonzero classical quark mass , where @xmath267 .
we now calculate the correction to the energy obtained by considering @xmath228 a small perturbation to the classical lagrangian @xmath268 .
the lagrangian , @xmath269 depends on only one degree of freedom , the rotation angle @xmath270 , through its time derivative @xmath271 , to first order in @xmath228 , the correction to the energy is minus the correction to the lagrangian @xcite , @xmath272 \bigg|_{\omega = \omega(j ) } \ , , \label{e of j}\ ] ] where @xmath8 is given as a function of @xmath5 through the classical relation ( [ j equation ] ) .
the correction ( [ e of j ] ) to the energy of the meson gives a correction to the slope ( [ classical regge eqn ] ) of a regge trajectory , @xmath273 using ( [ classical regge eqn ] ) for @xmath274 and ( [ l fluc calculated ] ) for @xmath228 , we obtain @xmath275 - \frac{4}{3\pi^2\sigma } \gamma^{-3 } + \frac{7}{6 \pi \sigma r e } + o\left(\gamma^{-5 } , \frac{1}{re\gamma}\right ) \ , .
\label{regge corrected}\ ] ] we write @xmath2 and @xmath265 as functions of @xmath4 using the definition ( [ e equation ] ) of @xmath4 and the classical equation of motion ( [ boundary condition ] ) . because @xmath2 and @xmath265 only appear in the small correction terms in the result ( [ regge corrected ] ) , we only need their leading order dependence on @xmath4 , @xmath276 substituting ( [ r , gamma of e ] ) in ( [ regge corrected ] ) gives @xmath277 - \frac{4}{3\sigma } \sqrt{\frac{m^3 e}{\pi } } + \frac{7}{12 } + o\left ( e^{-1/2 } \right ) \ , .
\label{j result}\ ] ] the leading term is the classical regge formula .
the next term is the leading correction due to string fluctuations .
the third term is a nonzero quark mass correction .
the fourth term is another correction due to string fluctuations .
( [ j result ] ) gives a meson regge trajectory @xmath6 .
[ cols=">,^ " , ] we used values @xmath278 and @xmath279 obtained from the cornell fits of heavy quark potentials @xcite .
this gives @xmath280 .
the only other parameter is the quark mass @xmath281 . in figure
[ regge figure ] , we plot @xmath5 versus the square of the energy ( [ e of j ] ) for quark masses of 30 mev , 100 mev , and 300 mev . for comparison
, we also plot the classical formula @xmath282 .
the points @xcite plotted on the graph are the @xmath283 , @xmath284 , @xmath285 , @xmath286 , @xmath287 , and @xmath288 mesons .
x s are the @xmath289 , @xmath119 , @xmath290 , @xmath291 , and @xmath292 .
we have added one to the value of the angular momentum @xmath5 in figure [ regge figure ] to account for the contribution of the spin of the quarks .
we have chosen a range of values for the quark masses in fig .
[ regge figure ] in order to give a qualitative picture of the dependence of the regge trajectory on the quark mass .
since ( [ j result ] ) does not include the contribution of quark fluctuations to the regge trajectory , this formula is incomplete .
we are now in the process of including the quark degrees of freedom in the functional integral ( [ param measure ] ) .
the boundary @xmath15 of the sheet @xmath9 becomes dynamical , and couples to the string fluctuations .
it is clear that a calculation of the contribution of these degrees of freedom is essential to understanding why the classical formula for regge trajectories works so well .
the primary results of the paper are ( [ param measure ] ) and ( [ j result ] ) .
we have expressed the path integral @xmath32 $ ] ( [ originalpartition ] ) of a renormalizable quantum field theory having classical vortex solutions as the path integral formulation of an effective string theory of vortices ( [ param measure ] ) .
this theory describes the two transverse fluctuations of the vortex at scales larger than the inverse mass of the lightest particle in the field theory .
our method is applicable to any field theory containing vortex solutions . using the string representation of @xmath32 $ ]
, we carried out a semiclassical expansion of the effective action @xmath141 $ ] about a classical solution of the effective string theory .
we calculated the contribution of these string fluctuations explicitly for the case where the worldline @xmath15 is generated by the trajectory of a quark
antiquark pair separated by a distance @xmath2 .
we are now calculating the contribution to the effective action @xmath141 $ ] due to the quantum fluctuations of the boundary .
we would like to thank n. brambilla for very helpful conversations .
this work was supported in part by the u. s. department of energy grant de - fg03 - 96er40956 .
we describe the string worldsheet by the function @xmath18 of the coordinates @xmath50 .
the physics of the vortex should be independent of the coordinate system we choose , so we require the theory to be invariant under a reparameterization of the coordinates , @xmath293 .
the tangent vectors to the vortex worldsheet are defined by taking derivatives of @xmath18 , @xmath294 where @xmath295 is a partial derivative with respect to one of the vortex coordinates .
the induced metric on the worldsheet @xmath18 is @xmath296 it is also convenient to define the square root of the determinant of the metric , @xmath297 we use the @xmath298 to define an antisymmetric tensor which describes the orientation of the string worldsheet , @xmath299 this quantity was defined by polyakov @xcite .
it is the projection of the antisymmetric tensor @xmath300 into the space of four dimensional tensors .
this tensor defines the orientation of the two dimensional vortex worldsheet in four space .
the quantity ( [ t mu nu def ] ) is also independent of the coordinate parameterization of the worldsheet @xmath18 .
we now describe the curvature of the vortex worldsheet .
we do this by taking covariant derivatives of the tangent vectors .
the covariant derivative of @xmath298 is @xmath301 where the @xmath302 are christoffel symbols , @xmath303 the covariant derivatives of the tangent vectors are orthogonal to the worldsheet , @xmath304 this identity is derived using the definition ( [ gamma def ] ) of the christoffel symbols , the definition ( [ metric appendix def ] ) of the metric , and the relationship between derivatives of different @xmath298 , @xmath305 the covariant derivatives of the tangent vectors are normal to the string worldsheet .
we therefore define a basis of normal vectors @xmath306 , which satisfy the conditions @xmath307 the @xmath306 are an orthonormal basis for the vectors normal to the worldsheet .
( [ no tangent covariant ] ) implies that @xmath308 for some tensor @xmath101 .
the tensor @xmath101 is called the extrinsic curvature tensor of the string worldsheet . with the definition ( [ curvature def ] ) , the curvature tensor @xmath101 is @xmath309 it is symmetric in the indices @xmath31 and @xmath310 due to the relationship ( [ tangent vector derivatives ] ) between derivatives of tangent vectors .
the extrinsic curvature of the string worldsheet can also be described using derivatives of the normal vectors .
the orthogonality of the @xmath311 and the @xmath306 implies @xmath312 therefore , the derivatives of the normal vectors can be written as @xmath313 the tensor @xmath314 is called the torsion , and it describes the twisting of the basis of normal vectors as we move along the worldsheet .
the torsion depends on our choice of the @xmath306 , so we will choose them so that the torsion is zero .
this is done by requiring that the @xmath306 satisfy the differential equation @xmath315 the equation ( [ torsion free ] ) is equivalent to the statement @xmath316 .
it is consistent with the conditions ( [ n def ] ) which define the normal vectors .
as long as the normal vectors have an orthonormal basis at one point , eq .
( [ torsion free ] ) guarantees they will be orthonormal in a neighborhood of that point . therefore , it is always possible to find a local , orthonormal , torsion free basis for the normal vectors .
there is one additional property of the normal vectors we will use .
the antisymmetric combination of the normal vectors is ( with proper ordering ) equal to the dual of the worldsheet orientation tensor @xmath317 ( [ t mu nu def ] ) , @xmath318 where @xmath319 the relationship ( [ dual normal ] ) can be understood by noting that any antisymmetric tensor is of the form @xmath320 the tensor @xmath321 is orthogonal to the @xmath298 , so it must be proportional to @xmath322 . squaring both of these tensors
gives @xmath323 therefore , @xmath321 and @xmath322 are equal up to an overall sign , which is fixed by choosing an appropriate ordering for the normal vectors .
the jacobian @xmath66 ( [ j_perp expressed ] ) is a faddeev
popov determinant , because fixing the position of the string in the field integrals is analogous to fixing a gauge in a gauge theory . in the string action
, we fix the degrees of freedom which generate the transformation @xmath324 which displaces the vortex . the jacobian @xmath66 is analogous to the faddeev
popov determinant in a gauge theory . in a gauge theory , where the @xmath44 function fixes the symmetry generated by the transformation @xmath325 the faddeev
popov determinant appears as a normalization for the @xmath44 function @xcite , @xmath326 \delta_{fp } e^{-s } \ , , \label{gauge part}\ ] ] where @xmath327 \,.\ ] ] the wilson loop ( [ gauge part ] ) is analogous to our equation ( [ seff ] ) for the effective action .
the determinant @xmath328 is analogous to @xmath66 . in the gauge theory , the faddeev
popov method is used to remove nonphysical degrees of freedom from the problem .
the @xmath44 function is inserted in the path integral eqnlessrefgauge part to fix the fields in some particular gauge .
this creates an integral over all gauges which appears as a normalization factor , and is removed . the @xmath44 function in eq .
( [ seff ] ) , on the other hand , fixes the position of the vortex sheet , which is a physical degree of freedom .
we want to evaluate the term , @xmath329 \,,\ ] ] in the effective action for a general string worldsheet .
we work in coordinates @xmath190 and @xmath189 , such that @xmath189 is the time in the lab frame , and @xmath190 is orthogonal to @xmath189 ( @xmath202 ) . in these coordinates ,
the functional integral ( [ w_1 int ] ) for @xmath198 takes the form @xmath330 \right\}\,.\ ] ] we consider the case where the metric is independent of @xmath189 , and we make the coordinate transformation @xmath210 defined by @xmath331 the coordinate @xmath212 runs from @xmath213 to @xmath214 , @xmath332 in these coordinates , the length of the string is @xmath216 , the proper length of the string , @xmath333 in the coordinate system @xmath217 , the metric is conformal , @xmath334 and @xmath335 \right\ } \ , .
\label{w_1 of x and t}\ ] ] we evaluate ( [ w_1 of x and t ] ) in a manner analogous to our treatment of @xmath226 in section [ w_2 section ] .
we fourier transform in both space and time , introducing variables @xmath222 and @xmath336 .
this transformation puts the action in ( [ w_1 of x and t ] ) in a diagonal form .
doing the @xmath337 and @xmath338 integrals gives @xmath339 \ , , \label{w_1 sum}\ ] ] where we have wick rotated @xmath224 to avoid the poles at @xmath340 .
the length @xmath207 is just the classical string energy @xmath341 divided by the string tension @xmath86 , since @xmath341 is @xmath342 the quantity @xmath343 is the length of the string measured in local co - moving coordinates , which are at rest with respect to the string .
this is different from the string length @xmath2 in the laboratory frame
. we will regulate @xmath194 using the results of lsw .
their result for @xmath174 is the following , @xmath344 where @xmath345 is the number of dimensions , @xmath346 is the area of the string worldsheet , @xmath347 is the length of its boundary , and @xmath182 is the scalar curvature of the sheet .
the @xmath176 are regulator masses , and the @xmath177 are appropriate coefficients .
the final term , @xmath179 , is finite in the limit where the @xmath348 .
lsw evaluate the finite term @xmath179 only for a straight string of length @xmath2 with fixed ends . in this case , the area of the sheet @xmath349 , the length of the boundary @xmath350 , and the curvature of the sheet is zero .
they then obtained the explicit contribution to the heavy quark potential : @xmath351 the first term in ( [ luscher reg ] ) renormalizes the string tension .
the second renormalizes the quark mass .
the third is the well known lscher term in the heavy quark potential .
since the extrinsic curvature vanishes for a flat sheet , we can identify the result ( [ luscher reg ] ) with our expression ( [ w_1 sum ] ) for @xmath194 , with @xmath207 replaced by @xmath2 : @xmath352 = \frac{1}{2\pi } r \sum_j \epsilon_j { \cal m}_j^2 \ln { \cal m}_j^2 + \sum_j \epsilon_j { \cal m}_j - \frac{\pi}{12 r } \ , .
\label{flat eigenvalue sum}\ ] ] eq .
( [ flat eigenvalue sum ] ) tells us how to regulate @xmath194 . replacing @xmath2 by @xmath207 in ( [ flat eigenvalue sum ] ) gives the regulated form of @xmath194 : @xmath353 the first term in ( [ modified luscher ] ) is still a string tension renormalization , since both the string tension contribution to the energy ( [ e sigma ] ) and the first term in eq .
( [ modified luscher ] ) are proportional to @xmath207 , the second term in eq .
( [ modified luscher ] ) is , as before , a renormalization of the quark mass .
the finite part of the contribution of @xmath194 to the action is @xmath354 this is the result stated in section [ regularization section ] . the result ( [ luscher r_p appendix ] ) is the lscher term , with the distance @xmath2 between the quarks replaced by the proper length @xmath207 of the string . our result is ( [ r_p appendix ] ) , the derivation of @xmath207 .
in this appendix , we show that the divergent part of @xmath226 for a general sheet is proportional to the integral of the scalar curvature @xmath182
. this agrees with the logarithmic divergence ( [ luscher reg ] ) derived by lsw by other means . to obtain the divergent part of @xmath226
, we carry out a fourier transform with respect to the variable @xmath212 . for functions defined on the interval @xmath355
, the @xmath44 function can be expressed as a sum of sines , @xmath356 where @xmath336 .
the fourier transform of an operator of the form @xmath357 can then be written @xmath358 using the formula ( [ fourier transform ] ) to evaluate ( [ w_2 given ] ) gives @xmath359 \ , , \label{fourier transformed}\end{aligned}\ ] ] the trace is over indices @xmath360 which run from 1 to 2 , and indices @xmath361 which run from 1 to @xmath236 .
the trace is cutoff at @xmath362 , the mass of the vector particle in the original field theory .
we expand ( [ fourier transformed ] ) for large @xmath363 and obtain the cutoff dependent part of @xmath226 , @xmath364 the term @xmath365 is equal to minus the scalar curvature @xmath182 , @xmath366 since the equation of motion ( [ class string simple ] ) implies @xmath367 .
the cutoff dependent part of @xmath226 is therefore @xmath368 eq .
( [ w_2 divergence ] ) agrees with the result of lsw for the leading semiclassical logarithmic divergence .
we want to evaluate @xmath226 , @xmath369 for the fluctuations about a straight string of length @xmath2 rotating with angular velocity @xmath8 . to evaluate this , we must determine the value of the extrinsic curvature @xmath101 .
the definition of the extrinsic curvature is @xmath102 the string @xmath370 is @xmath371 the @xmath372 are a basis of orthonormal unit vectors in minkowski space .
the @xmath306 are a basis for the vectors normal to @xmath370 .
we choose the basis @xmath373 with this choice for the @xmath306 , @xmath374 is zero , because the @xmath375 component of @xmath370 is zero . the only nonzero component of @xmath376 is @xmath377 now that we know what @xmath101 is , we can evaluate ( [ v trace ] ) . inserting ( [ nonzero k ] ) into ( [ v trace ] )
gives @xmath378 the traces in ( [ final rotating potential ] ) are defined as sums over the eigenvalues of the given operators .
replacing the traces with explicit sums gives @xmath379 where @xmath380 the eigenvalues @xmath381 are determined by the eigenfunction equation @xmath382 with the boundary conditions @xmath383 .
the difference between the traces in eq .
( [ final rotating potential ] ) is logarithmically dependent on the cutoff @xmath76 ( the mass of the dual gluon ) .
( [ eigenfunction equation ] ) has the form of the schredinger equation , with the potential @xmath384 .
this potential is an analytic continuation of the potential @xmath385 , whose eigenfunctions can be expressed terms of hypergeometric functions @xcite . using this result , we find the eigenfunctions @xmath386 the eigenvalues @xmath381 are @xmath387 where @xmath388 satisfies the transcendental equation @xmath389 and @xmath390 . there is no @xmath391 eigenvalue , despite the fact that @xmath392 satisfies ( [ alpha def ] ) , because the corresponding eigenvalue @xmath393 makes @xmath394 zero everywhere .
we will carry out the sum ( [ sum to do ] ) , @xmath395 by converting it to a contour integral .
we will find a function @xmath396 which has zeros whenever @xmath397 .
we will find another function @xmath398 which has zeros whenever @xmath399 .
we will then define a function @xmath400 , @xmath401 the function @xmath400 has poles of residue @xmath402 when @xmath403 and poles of residue @xmath404 when @xmath399 .
we then rewrite the sum ( [ w_2 sum before contour ] ) as a contour integral , @xmath405 the contour of the integral ( [ w_2 contour abstract ] ) lies along the imaginary axis , and on a semicircle at @xmath406 with the real part of @xmath407 positive . to write @xmath226 as the contour integral ( [ w_2 contour abstract ] )
, we need to find the functions @xmath396 and @xmath398 .
the function @xmath398 is @xmath408 which is zero for @xmath399 .
we find @xmath396 by recalling that the eigenfunctions ( [ eigenfunctions ] ) vanish at @xmath409 .
therefore , @xmath410 has zeros at @xmath411 for @xmath412 odd , and @xmath413 has zeros at @xmath411 for @xmath412 even .
thus , @xmath414 the factor @xmath415 removes nonphysical zeros which appear because @xmath416 .
these zeros correspond to the @xmath391 `` eigenfunction '' which is zero everywhere for @xmath393 .
the function @xmath400 is @xmath417 \ , .
\label{f int found}\ ] ] inserting ( [ f int found ] ) in ( [ w_2 contour abstract ] ) and integrating by parts gives @xmath418 \ , .
\label{contour defined}\ ] ] now , instead of having poles at @xmath411 and @xmath419 , the integrand has branch points at these points .
the branch cuts run from @xmath420 to @xmath421 for all @xmath412 along the real axis .
there is one branch cut for each value of @xmath412 .
since the contour either includes both @xmath420 and @xmath421 or excludes both these points , none of these branch cuts cross the contour of integration .
there are no branch points at @xmath422 , since both the numerator and denominator vanish there .
none of the branch cuts crosses the contour , so the contour is still closed , and the integration by parts does not produce a boundary term . the contour of the integral ( [ contour defined ] ) lies on the imaginary axis and a semicircle passing through positive real infinity .
we rewrite ( [ contour defined ] ) as two integrals over the different pieces of the contour . for large values of the cutoff @xmath76 , the action @xmath226 is @xmath423 \nonumber \\ & & - t \frac{1}{4\pi } \int_{-\pi/2}^{\pi/2 } d\theta \lambda e^{i\theta } \left[-2\frac{\omega}{\lambda } e^{-i\theta } \tan\left(\frac{r_p}{2 } \omega\right ) \cot\left(r_p \lambda e^{i\theta}\right ) + o(\lambda^{-2 } )
\right ] \,.\end{aligned}\ ] ] for large @xmath76 , @xmath424 is proportional to the sign of @xmath270 , so the @xmath270 integral vanishes .
the @xmath425 integral is symmetric under @xmath426 .
changing variables to @xmath427 gives @xmath428 \ , .
\label{s integral}\ ] ] the numerator of in ( [ s integral ] ) is approximately @xmath429 for large values of @xmath430 .
we use this fact to extract the divergent part of ( [ s integral ] ) , getting @xmath431 - t \frac{v^2 \gamma}{\pi r } f\left ( v \right ) \ , .
\label{s_2 lambda cutoff}\ ] ] we have replaced @xmath207 with its definition @xmath432 @xmath433 is the quark boost factor .
the function @xmath250 contains the rest of the integral , @xmath434 \,.\ ] ] for @xmath256 , the asymptotic value of @xmath250 is @xmath435 the cutoff @xmath76 used in ( [ s_2 lambda cutoff ] ) is the cutoff in the @xmath212 coordinate .
we must express @xmath76 in terms of the cutoff @xmath77 for the @xmath190 coordinate , which measures physical distance .
the cutoffs @xmath76 and @xmath77 are related by the equation @xmath436 , or @xmath437 inserting ( [ lambda of m ] ) into ( [ s_2 lambda cutoff ] ) gives @xmath438 - t \frac{v^2 \gamma}{\pi r } f\left ( v \right ) \,.\ ] ] using the classical equation of motion ( [ boundary condition ] ) then gives @xmath439 - t \frac{v^2 \gamma}{\pi r } f\left ( v \right ) \ , .
\label{s_2 final}\ ] ] | starting from a field theory containing classical vortex solutions , we obtain an effective string theory of these vortices as a path integral over the two transverse degrees of freedom of the string .
we carry out a semiclassical expansion of this effective theory , and use it to obtain corrections to regge trajectories due to string fluctuations . |
the hera collider facility in hamburg , germany , was a unique machine for lepton - proton scattering at highest energies .
for the two experiments h1 and zeus protons with an energy of up to @xmath2 gev and electrons or positrons with an energy of @xmath3 gev were collided .
this is equivalent to a maximal centre of mass energy of @xmath4 gev . at the end of june 2007
the data taking finished . in deep inelastic scattering ( dis ) of leptons off nucleons the substructure of the nucleons
was discovered .
dis continues to be the tool for exploring the substructure of the nucleons with high precision , i.e. measuring their quark and gluon content in the form of so called parton distribution functions ( pdfs ) .
the evolution of pdfs is a sensitive test of our understanding of qcd dynamics , which is expressed in the form of pdf evolution equations .
furthermore a precise knowledge of pdfs is vital for measurements at hadron colliders , such as the lhc .
the kinematics of the scattering are described in terms of the lorentz invariant quantities : the bjorken scaling variable @xmath5 , the inelasticity @xmath6 , and the virtuality @xmath7 .
the hera experiments have made measurements of the proton structure for @xmath7 values of up to @xmath8gev@xmath9 and @xmath5 values down to @xmath10 .
one of the most fundamental measurements to be performed is that of the neutral current ( nc ) inclusive cross section for the reaction @xmath11 , which can be expressed in the form @xmath12 with @xmath13 and the structure functions @xmath14 , @xmath15 and @xmath16 . at leading order
, the structure functions relate to the pdfs as @xmath17 additional information can be obtained from charged current ( cc ) cross section for the reaction @xmath18 : @xmath19 the gluon density @xmath20 , which does not enter the inclusive cross section calculations at leading order , is constrained by the @xmath14 scaling violations , jet cross sections and the measurement of @xmath15 .
both zeus and h1 have performed inclusive cross section measurements with comparable precision and covering a similar kinematic range . therefore the results of the two experiments may be combined to provide a single set containing all inclusive hera cross sections and to be used for further qcd analyzes .
the averaging procedure , described in detail in @xcite , is without theoretical assumptions .
only small theoretical corrections are needed for swimming points to a common @xmath21 grid and for adjusting earlier measurements performed at a proton beam energy of @xmath22 gev .
the precision cross section data from the hera experiments are typically reported with three different components of the measurement uncertainty : statistical and systematic uncertainties , where the latter consist of parts uncorrelated or correlated between different kinematic domains .
the correlated uncertainties , which are due to effects like the luminosity measurement or shifts in the electron energy calibration need to be treated correctly in the averaging procedure . as a result , not only the uncorrelated and statistical part of the uncertainties are reduced .
because the measurements by the two experiments have different sensitivities to the correlated uncertainties , these can be constrained further and thus reduced .
the combination procedure is performed separately for the sets of @xmath23 and @xmath24 scattering and using both nc and cc data @xcite .
the original data and the combination are shown for a few exemplary values of @xmath5 of the @xmath23 nc data set in figure [ fig : h1zcombfit ] , left .
a remarkable reduction of the uncertainties is apparent .
figure [ fig : h1zcombfit ] , right , shows the full @xmath23 data set , which is seen to cover four orders of magnitude in @xmath5 and @xmath7 with very high precision .
the compatibility of the two experiments is observed to be very good with a total @xmath25 for @xmath26 averaged points .
the combined data is in the following used to perform a qcd analysis at nlo using hera data only . for the analysis , which is called herapdf @xmath1 ,
the dglap evolution equations are used to evolve the pdfs , which are parametrized at the starting scale of @xmath27gev@xmath9 .
a full evaluation of experimental and theory model uncertainties is performed .
the resulting pdfs are shown in figure [ fig : qcdfits ] at a scale of @xmath28gev@xmath9 and compared to earlier separate analyzes of the h1 and zeus collaborations @xcite .
a reduction of the uncertainties is seen , which is due to the improved combined data used as input .
this is most notable on the gluon density , which dominates all other partons at low @xmath5 whose uncertainty is typically also larger than the quark pdfs , which are directly accessed in dis .
while data taking at hera has stopped , more precise measurements have become available recently or are expected for the near future . in general , analyzes at low and intermediate @xmath29gev@xmath9 can be performed using relatively small data sets .
therefore the new high precision analyzes of h1 @xcite were performed using data from the hera - i running period , which ended in the year 2000 .
the total measurement uncertainties were reduced to as low as @xmath30 per point , which is about a factor of two better than previous data covering this kinematic domain @xcite . as an example , the new results on the structure function @xmath14 from @xcite are shown in figure [ fig : newdata ] , left . the analyzes at higher @xmath31gev@xmath9 are in general more constrained by the available statistics .
therefore improvements are expected from the analysis of the hera - ii data , where nearly @xmath32pb@xmath33 were collected per experiment after the hera luminosity upgrade starting from the year 2003 .
the data sample is about balanced between @xmath23 and @xmath24 .
a new feature of the hera - ii data is also beam polarization , where typical average values of @xmath34 are reached .
zeus has recently published new results on the inclusive nc and cc @xmath24 cross section @xcite .
figure [ fig : newdata ] , right , shows the cc cross sections and highlights their dependence on different quark flavours in the proton .
these are the first double differential inclusive measurements using the polarized electron beam .
also the employed @xmath24 hera - ii data samples have an integrated luminosity ten times larger than the previously used hera - i sample .
the hera experiments measure the proton structure in a wide range of @xmath5 and @xmath7 , thus providing stringent tests of qcd and valuable input for the lhc .
a new procedure combines the already published data to obtain one hera data set with improved uncertainties .
a combined qcd fit , herapdf 0.1 , is performed using this combined data set .
it is able to describe the data and provides pdfs with much reduced uncertainties compared to previous fits performed separately by the h1 and zeus collaborations .
in addition new measurements are available with significantly increased experimental precision .
99 c. adloff _ et al . _
[ h1 collaboration ] , eur .
j. c * 30 * ( 2003 ) 1 [ arxiv : hep - ex/0304003 ] .
s. chekanov _ et al . _
[ zeus collaboration ] , eur .
j. c * 42 * ( 2005 ) 1 [ arxiv : hep - ph/0503274 ] .
[ h1 collaboration ] , arxiv:0904.3513 [ hep - ex ] . c. adloff _ et al . _
[ h1 collaboration ] , eur .
j. c * 21 * ( 2001 ) 33 [ arxiv : hep - ex/0012053 ] .
s. chekanov _ et al . _
[ zeus collaboration ] , eur .
j. c * 21 * ( 2001 ) 443 [ arxiv : hep - ex/0105090 ] .
s. chekanov _ et al . _
[ zeus collaboration ] , arxiv:0901.2385 [ hep - ex ] .
s. chekanov _ et al . _
[ zeus collaboration ] , arxiv:0812.4620 [ hep - ex ] . | new and previously published measurements on the deep inelastic @xmath0 scattering cross section by the h1 and zeus collaborations are presented .
the uncertainties can be significantly reduced by a model independent combination procedure , which treats the systematic error correlations in a coherent way .
the combined h1 and zeus measurements of the inclusive neutral and charged current cross sections are used to perform a common nlo qcd fit , called herapdf @xmath1 .
the resulting set of parton density functions ( pdfs ) have a much improved experimental uncertainty compared to previous extractions using the uncombined h1 or zeus data . |
the optics community has taken a lot of effort into the development of surface modeling methods for optical design @xcite . unfortunately
, none of those methods are developed to improve manufacturability @xcite or to reduce the difficulty of fabrication of optical surfaces .
for example , zernike polynomials are widely applied to describe surface errors and express wavefront data in the field of optical fabrication and testing . because there is a theoretical relationship between zernike coefficients and seidel aberrations often observed in optical tests @xcite . however , basically , the standard 36-term zernike polynomial set does not aim to or help to optimize the process of optical fabrication or polishing .
when it comes to ultra - precision fabrication of optical surfaces , ion beam technologies play a key role in improving surface precision to extreme .
as one of the most precise methods , ion beam figuring ( ibf ) @xcite is employed for finishing lithography optics and telescope mirrors .
ibf is realized with a dwell time algorithm akin to that used in computer controlled optical surfacing ( ccos ) .
longer time the ion beam dwelling at a point results in more material removal around the point , thus it allows correction of surface figures through controlled variations of scanning velocity . in particular , one - dimensional ibf @xcite , or sometimes called ion beam profiling @xcite , applies to the fabrication of elongated synchrotron optics such as aspherical x - ray mirrors .
compared with the conventional ibf , one - dimensional ibf is generally implemented by a simpler algorithm meanwhile the long - rectangular - shaped ion beam adopted in one - dimensional case leads to higher output efficiency under the condition for ensuring similar finishing precision .
inspired by zernike polynomials bridging the relationship between surface errors and optical aberrations , we propose a surface modeling method on the basis of double fourier series @xcite to narrow the gap between optical fabrication and surface characterization .
the fourier series decomposition of an optical surface produces a set of wave surfaces with a sinusoidal profile , which by and large are of different periods , amplitudes , and propagation directions .
conceptually , we can sequentially fabricate the decomposed wave surfaces by linear scanning with a linear ion source and the superposition of those fabricated wave surfaces finally build up a surface that approximates to the desired optical surface .
moreover , the power spectral density ( psd ) analysis is the statistical analysis of the spatial - distributed wave surfaces described as components of double fourier series . as to the characterization of optical surfaces , the psd function
is typically utilized @xcite , especially for evaluating errors in the mid - spatial frequency ( msf ) range .
so , we realized , a surface descriptor based on fourier series can build a relationship between the technique of optical fabrication with linear ion source and the psd - based surface characterization method .
this paper introduces the basic theory of surface modeling and decomposition , then illustrates the principle of optical fabrication with linear ion source , and finally demonstrates two applications , i.e. , fabrication of large - aperture beam sampling gratings ( bsgs ) and removal of errors in the msf range .
a continuous smoothing surface @xmath0 can be described as @xmath1 where @xmath0 is periodic by @xmath2 in x direction and @xmath3 in y direction ; @xmath4 , @xmath5 ; @xmath6 is a complex number . in particular
, @xmath0 is a real function that is conjugate symmetric thus the imaginary part can be omitted . after adding a proper piston to every component , @xmath7
, we get a slightly elevated surface , @xmath8 , that is suitable for indicating the spatial distribution of material removals and is displayed as @xmath9 where @xmath10 represents the piston or extra material removal ; note that the arbitrary component @xmath11 . since ion etching is a subtractive process rather than an additive process , the desired material removal should be nonnegative thus it requires that @xmath12 .
continued by equation [ eq : removal_surface1 ] , we put forward a new formula to decrease the number of scan strokes and improve the reachability of optical surfacing , which is represented as @xmath13 where @xmath14 and @xmath15 not equal to @xmath16 simultaneously ; @xmath17 denotes a small positive quantity and represents the compensated material removal .
since there is an optimum removal in ion - beam figuring @xcite , @xmath17 usually is not infinitesimal in order to make sure that the scan speed is properly limited .
the component @xmath18 , indicating the decomposed distribution of material removals , is shown as a sinusoidal wave surface , where the propagation direction is @xmath19 or @xmath20 and the period is @xmath21 . @xmath22 where @xmath23 represents a material removal map by m times n , which is the discrete representation of @xmath24 ; two - dimensional fast fourier transform ( denoted as fft2 in this paper ) algorithm is used for calculating the matrix of fourier coefficients , @xmath25 .
to exclude the dc component , @xmath26 , it needs to assign that : @xmath27 .
the material removal map can be effectively approximated by the symmetric rectangular partial sum of double fourier series @xcite .
the ( @xmath28)-th symmetric rectangular partial sum is defined as @xmath29 then we have the truncated errors @xmath30 to evaluate the feasibility of the approximation , we use psd function to analyze the truncated errors referring to as surface errors . the psd analysis of the truncated errors can be implemented by the two - dimensional fft algorithm , which is displayed as follows : @xmath31
reactive ion etching ( rie ) is a useful method in the fabrication of diffraction optics .
recently we fabricated large - aperture bsgs by the rie process , where the ion beam emitted from linear ion source scans over bsgs along a linear guide meanwhile the rotary leaf scans along the elongated footprint of ion beam , thus the spatial distribution of etch depths is finely adjusted for improving the diffractive uniformity of bsg @xcite .
now we try to exploit more applications by combining optical fabrication technique with linear ion source and the aforementioned surface modeling method . fig .
[ fig : optsurf ] illustrates the basic principle of optical fabrication with linear ion source and also serves as a schematic of the fabrication of bsgs .
the two shutters can collimate the ion beam emitted from the linear ion source and adjust the width of ion beam spot ( see fig . [
fig : optsurf ] ) projected on the workpiece or bsg substrate .
the workpiece mounted on a workbench spins about its center normal line and moves in the transverse horizontal direction in accompany with the workbench .
that means the ion beam having an elongated rectangular footprint can scan over the workpiece along the arbitrary direction on the workpiece surface .
linear ion sources , or ion sources with a large - aperture rectangular grid , are suitable for ion figuring of optical surfaces with mainly a one - dimensional profile such as synchrotron reflective mirrors .
fortunately , we only need to fabricate wave surfaces with a one - dimensional sinusoidal profile with the help of surface decomposition .
a set of decomposed wave surfaces linearly build up a superposed surface approximating to the desired surface .
moreover , we have compared two approach in our previous work @xcite to calculate the dwell time distribution in corresponding with a sinusoidal profile .
in the fabrication of large - aperture bsgs , the spatial distribution of etch depths should be finely adjusted to improve the diffractive uniformity of bsg . the mesh graph on the top right of fig .
[ fig : reduce ] is a typical etch depth map , which describes the desired etch depths spatially distributed on the bsg substrate .
the tolerance analysis indicates that the rms of etch depth differences should be within 0.4 nm , which means the rms of truncated errors , @xmath32 , is not greater than 0.4 nm . as is shown on the top left of fig .
[ fig : reduce ] , the psd in the low - spatial frequency range is significantly higher than that in the mid / high - spatial frequency range , which indicates that the first several components or decomposed wave surfaces mainly contribute to the surface form .
after psd analysis , as is shown in fig .
[ fig : decomp ] , the first 5 sinusoidal wave surfaces including @xmath33 are selected and the summation ( @xmath34 ) of the 5 components is well approximated to the desired etch depth map . the etch depth differences , obtained by removing @xmath34 from the desired etch depth map , is shown on the lower part of fig .
[ fig : reduce ] , where the peak - to - valley value is 0.19 nm and the average of etch depth differences is 15.95 nm .
( see fig .
[ fig : decomp ] ) from the etch depth map . ] ) and their summation ( @xmath35 ) . ] for this example , to fabricate a bsg of high diffractive uniformity , it needs to carry out at least 6 times of linear scanning of ion beam . to start with removing a uniform layer of thickness of 15.95 nm
, we then fabricate the 5 wave surfaces sequentially by linear scanning of ion beam . as is shown in fig .
[ fig : decomp ] , for the 5 sinusoidal wave surfaces , the rotary directions are ( 0 , 1 ) , ( 1 , -1 ) , ( 1 , 0 ) , ( 1 , 1 ) , and ( 2 , 0 ) ; the periods are 400 mm , 283 mm , 400 mm , 283 mm , and 100 mm . in the process of ion beam etching
, the bsg should be rotated according to the rotary directions and the width of ion beam spot should be less than a half of the period of wave surface . as is demonstrated above , optical fabrication with linear ion source can be used to finely adjust etch depths of large - aperture diffractive optics based on the concept of surface decomposition .
we prepared a large - aperture optical surface ( @xmath36 ) polished by ccos with small tools .
the measured surface map is shown on the upper right of fig .
[ fig : errors ] . and
more msf errors are found in the horizontal direction than those in the vertical direction , which is attributed to the raster scan path adopted in the small - tool polishing process .
the corresponding psd plots are arranged on the left of fig .
[ fig : errors ] , where the solid curve and the dotted curve represent the psd functions along the horizontal and the vertical , respectively . as is shown on the upper left of fig .
[ fig : errors ] , there is a high peak at the solid psd curve in the spatial frequency of around @xmath37 . in general , a peak at the psd curve indicates more errors existing in the spatial frequency range around the peak . to remove errors in a specific range of spatial frequencies
, we should target the errors displayed as sinusoidal wave surfaces and then remove them one by one via optical surfacing with linear ion source . given that @xmath38 ; @xmath39 ; @xmath40 and provided that @xmath41 ( only the errors distributed along the horizontal are of interest )
, then we have @xmath42 , which refers to @xmath43 .
actually , to remove the peak totally , it requires a set of sinusoidal wave surfaces . fig .
[ fig : waves ] shows a set of sinusoidal wave surfaces , @xmath44 , stacking up to remove errors in the spatial frequency range around @xmath37 .
the summation , @xmath45 , represents the to - be - etched material removal , which is shown on the right of fig .
[ fig : waves ] . after reducing the to - be - etched material removal from the original surface map ,
we obtain a surface map with less errors in the frequency range nearby @xmath37 ( shown on the bottom right of fig .
[ fig : errors ] ) and the peak disappears in the corresponding psd plot . , summing up to remove errors in the spatial frequency range around @xmath37 . ]
the simulation result indicates that , theoretically , one - dimensional polishing is able to get rid of msf errors spatially distributed vertical to the scanning direction .
it also suggests that ultra - precision optical surfacing with linear ion source is a feasible and preferred approach to reduce errors in msf range . because optical surfacing with linear ion source is a superposition of multiple one - dimensional polishing processes and there is no discrete stitching of polishing spots in the course of polishing @xcite . as illustrated in fig .
[ fig : optsurf ] , the aperture of optical element is less than the length of rectangular beam spot thus the ion beam can fully cover and scan over the optical surface without stitching along the length direction .
we propose a surface modeling method based on double fourier series and the concept of surface decomposition makes it feasible to fabricate an ultraprecise optical surface with linear ion source . also , we demonstrate two applications .
the first application shows that optical fabrication with linear ion source can be applied to direct fabricating a bsg of high diffractive uniformity without post - processing ( for example , chemical mechanical polishing ) for correcting the spatial distribution of diffraction efficiency .
the second application presents a new approach to remove surface errors in msf range and we think the basic concept can be realized with other advanced polishing tools .
moreover , optical fabrication with linear ion source will significantly improve the working efficiency compared with the conventional ibf , though small round beam has a higher reachability in the fabrication of complex surfaces .
youth innovation promotion association of the chinese academy of sciences .
we thank mr . zhentong liu for the help in drawing the schematic of optical fabrication with linear ion source . | we present a concept of surface decomposition extended from double fourier series to nonnegative sinusoidal wave surfaces , on the basis of which linear ion sources apply to the ultra - precision fabrication of complex surfaces and diffractive optics .
it is the first time that we have a surface descriptor for building a relationship between the fabrication process of optical surfaces and the surface characterization based on psd analysis , which akin to zernike polynomials used for mapping the relationship between surface errors and seidel aberrations . also , we demonstrate that the one - dimensional scanning of linear ion source is applicable to the removal of surface errors caused by small - tool polishing in raster scan mode as well as the fabrication of beam sampling grating of high diffractive uniformity without a post - processing procedure .
the simulation results show that , in theory , optical fabrication with linear ion source is feasible and even of higher output efficiency compared with the conventional approach . 99 g. forbes , `` shape specification for axially symmetric optical surfaces , '' * 15*(8 ) , 52185226 ( 2007 ) .
p. jester , c. menke , and k. urban , `` b - spline representation of optical surfaces and its accuracy in a ray trace algorithm , '' * 50*(6 ) , 822828 ( 2011 ) .
p. jester , c. menke , and k. urban , `` wavelet methods for the representation , analysis and simulation of optical surfaces , '' i m a j. appl . math . * 77*(4 ) , 495515 ( 2012 ) . g. forbes , `` fitting freeform shapes with orthogonal bases , '' * 21*(16 ) , 1906119081 ( 2013 ) .
c. ferreira , j. l. lpez , r. navarro , and e. p. sinusa , `` orthogonal basis with a conicoid first mode for shape specification of optical surfaces , '' * 24*(5 ) , 54485462 ( 2016 ) .
g. forbes , `` manufacturability estimates for optical aspheres , '' * 19*(10 ) , 99239942 ( 2011 ) .
r. k. tyson , `` conversion of zernike aberration coefficients to seidel and higher - order power - series aberration coefficients , '' * 7*(6 ) , 262264 ( 1982 ) .
x. xie and s. li , `` ion beam figuring technology '' in _ handbook of manufacturing engineering and technology , _ andrew y. c. nee , ed .
( springer london , 2015 ) , pp . 13431390 .
l. zhou , m. idir , n. bouet , k. kaznatcheev , l. huang , m. vescovi , y. dai , and s. li , `` one - dimensional ion - beam figuring for grazing - incidence reflective optics , '' j. synchrotron radiat . *
23*(1 ) , 182186 ( 2016 ) .
l. peverini , i. kozhevnikov , a. rommeveaux , p. vaerenbergh , l. claustre , s. guillet , j .- y .
massonnat , e. ziegler , and j. susini , `` ion beam profiling of aspherical x - ray mirrors , '' nucl .
instr . meth .
phys .
res .
a * 616*(2 ) , 115118 ( 2010 ) .
f. mricz and d. waterman , `` convergence of double fourier series with coefficients of generalized bounded variation , '' j. math .
anal .
appl . * 140*(1 ) , 3449 ( 1989 ) .
r. n. youngworth , b. b. gallagher , and b. l. stamper , `` an overview of power spectral density ( psd ) calculations , '' proc .
spie * 5869 * , 58690u ( 2005 ) .
l. zhou , y. dai , x. xie , and s. li , `` optimum removal in ion - beam figuring , '' precis .
eng . * 34*(3 ) , 474479 ( 2010 ) .
l. wu , k. qiu , y. liu , and s. fu , `` algorithms for finely adjusting etch depths to improve the diffraction efficiency uniformity of large - aperture bsg , '' j. opt .
* 17*(3 ) , 035401 ( 2015 ) .
l. wu , k. qiu , and s. fu , `` fine - tuning the etch depth profile via dynamic shielding of ion beam , '' nucl .
instr . meth .
phys .
res .
b * 381 * , 72 - 75 ( 2016 ) .
l. wu , k. qiu , and s. fu , `` variable - spot ion beam figuring , '' nucl .
instr . meth .
phys .
res .
b * 370 * , 7985 ( 2016 ) .
s. li , c. jiao , x. xie , and l. zhou , `` stitching algorithm for ion beam figuring of optical mirrors , '' sci .
china , ser .
e * 52*(12 ) , 35803586 ( 2009 ) . |
this review deals with the analysis of influences that one system , be it physical , economical , biological or social , for example , can exert over another .
in several scientific fields , the finding of the influence network between different systems is crucial . as examples
, we can think of gene influence networks @xcite , relations between economical variables @xcite , communication between neurons or the flow of information between different brain regions @xcite , or the human influence on the earth climate @xcite , and many others .
the context studied in this report is illustrated in figure [ network : fig ] .
for a given system , we have at disposal a number of different measurements . in neuroscience
, these can be local field potentials recorded in the brain of an animal ; in solar physics , these can be solar indices measured by sensors onboard some satellite ; in the study of turbulent fluids , these can be the velocity measured at different scales in the fluid ( or can be as in the figure , the wavelet analysis of the velocity at different scales ) . for these different examples ,
the aim is to find dependencies between the different measurements , and if possible , to give a direction to the dependence . in neuroscience
, this will allow to understand how information flows between different areas of the brain ; in solar physics , this will allow to understand the links between indices and their influence on the total solar irradiance received on earth ; in the study of turbulence , this can confirm the directional cascade of energy from large down to small scales . in a graphical modeling approach
, each signal is associated to a particular node of a graph , and dependence are represented by edges , directed if a directional dependence exists .
the questions addressed in this paper concern the assessment of directional dependence between signals , and thus concern the inference problem of estimating the edge set in the graph of signals considered .
climatology and neuroscience were already given as examples by norbert wiener in 1956 @xcite , a paper which inspired econometrist clive granger to develop what is now termed granger causality @xcite .
wiener proposed in this paper that a signal @xmath0 causes another time series @xmath1 , if the past of @xmath0 has a strictly positive influence on the quality of prediction of @xmath1 .
let us quote wiener @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ as an application of this , let us consider the case where @xmath2 represents the temperature at 9 a.m. in boston and @xmath3 represents the temperature at the same time in albany .
we generally suppose that weather moves from west to east with the rotation of the earth ; the two quantities @xmath4 and its correlate in the other direction will enable us to make a precise statement containing some if this content and then verify whether this statement is true or not . or again , in the study of brain waves we may be able to obtain electroencephalograms more or less corresponding to electrical activity in different part of the brain . here
the study of coefficients of causality running both ways and of their analogues for sets of more than two functions @xmath5 may be useful in determining what part of the brain is driving what other part of the brain _ in its normal activity_. " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in a wide sense , granger causality can be summed up as a theoretical framework based on conditional independence to assess directional dependencies between time series . it is interesting to note that norbert wiener influenced granger causality , as well as another field dedicated to the analysis of dependencies : information theory .
information theory has led to the definition of quantities that measure the uncertainty on variables using probabilistic concepts .
furthermore , this has led to the definition of measures of dependence based on the decrease in uncertainty relating to one variable after observing another one .
usual information theory is , however , symmetrical .
for example , the well - known mutual information rate between two stationary time series is symmetrical under an exchange of the two signals : the mutual information assesses the undirectional dependence .
directional dependence analysis viewed as an information - theoretic problem requires the breaking of the usual symmetry of information theory .
this was realized in the 1960 s and early 1970 s by hans marko , a german professor of communication .
he developed the bidirectional information theory in the markov case @xcite .
this theory was later generalized by james massey and gerhard kramer , to what we may now call directed information theory @xcite .
_ it is the aim of this report to review the conceptual and theoretical links between granger causality and directed information theory . _
many information - theoretic tools have been designed for the practical implementation of granger causality ideas .
we will not show all of the different measures proposed , because they are almost always particular cases of the measures issued from directed information theory .
furthermore , some measures might have been proposed in different fields ( and/or at different periods of time ) and have received different names .
we will only consider the well - accepted names .
this is the case , for example , of ` transfer entropy ' , as coined by schreiber in 2000 @xcite , but which appeared earlier under different names , in different fields , and might be considered under slightly different hypotheses .
prior to developing a unified view of the links between granger causality and information theory , we will provide a survey of the literature , concentrating on studies where information theory and granger causality are jointly presented .
furthermore , we will not review any practical aspects , nor any detailed applications . in this spirit , this report is different from @xcite , which concentrated on the estimation of information quantities , and where the review is restricted to transfer entropy . for reviews on the analysis of dependencies between systems and for applications of granger causality in neuroscience
, we refer to @xcite .
we will mention however some important practical points in our conclusions , where we will also discuss some current and future directions of research in the field .
we will not debate the meaning of causality or causation .
we instead refer to @xcite .
however , we must emphasize that granger causality actually measures a statistical dependence between the past of a process and the present of another . in this respect ,
the word causality in granger causality takes on the usual meaning that a cause occurs _
prior _ its effect .
however , nothing in the definitions that we will recall precludes that signal @xmath0 can simultaneously be granger caused by @xmath1 and be a cause of @xmath1 ! this lies in the very close connection between granger causality and the feedback between times series .
granger causality is based on the usual concept of conditioning in probability theory , whereas approaches developed for example in @xcite relied on causal calculus and the concept of intervention . in this spirit , intervention is closer to experimental sciences , where we imagine that we can really , for example , freeze some system and measure the influence of this action on another process .
it is now well - known that causality in the sense of between random variables can be inferred unambiguously only in restricted cases , such as directed acyclic graph models @xcite . in the granger causality context , there is no such ambiguity and restriction . in his nobel prize lecture in 2003 ,
clive w. granger mentioned that in 1959 , denis gabor pointed out the work of wiener to him , as a hint to solve some of the difficulties he met in his work .
norbert wiener s paper is about the theory of prediction @xcite . at the end of his paper , wiener proposed that prediction theory could be used to define causality between time series .
granger further developed this idea , and came up with a definition of causality and testing procedures @xcite . in these studies
, the essential stones were laid .
granger s causality states that a cause must occur before the effect , and that causality is relative to the knowledge that is available .
this last statement deserves some comment .
when testing for causality of one variable on another , it is assumed that the cause has information about the effect that is unique to it ; _ i.e. _ this information is unknown to any other variable . obviously , this can not be verified for variables that are not known .
therefore , the conclusion drawn in a causal testing procedure is relative to the set of measurements that are available .
a conclusion reached based on a set of measurements can be altered if new measurements are taken into account .
mention of information theory is also present in the studies of granger . in the restricted case of two gaussian signals ,
granger already noted the link between what he called the ` causality indices ' and the mutual information ( eq .
5.4 in @xcite ) .
furthermore , he already foresaw the generalization to the multivariate case , as he wrote in the same paper : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` in the case of @xmath6 variables , similar equations exist if coherence is replaced by partial coherence , and a new concept of partial information is introduced . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ granger s paper in 1969 does not contain much new information , but rather , it gives a refined presentation of the concepts . during the 1970 s , some studies , _
@xcite , appeared that generalized along some of the directions granger s work , and related some of the applications to economics . in the early 1980 s
, several studies were published that established the now accepted definitions of granger causality @xcite .
these are natural extensions of the ideas built upon prediction , and they rely on conditional independence . finally , the recent studies of dalhaus and eichler allowed the definitions of granger causality graphs @xcite .
these studies provide a counterpart of graphical models of multivariate random variables to multivariable stochastic processes . in two studies published in 1982 and 1984 @xcite , geweke , another econometrician , set up a full treatment of granger causality testing for the gaussian case , which included the idea of feedback and instantaneous coupling . in @xcite ,
the study was restricted to the link between two time series ( possibly multidimensional ) . in this study ,
geweke defined an index of causality from @xmath0 to @xmath1 ; it is the logarithm of the _ ratio _ of the asymptotic mean square error when predicting @xmath1 from its past only , to the asymptotic mean square error when predicting @xmath1 from its past and from the past of @xmath0 .
geweke also defined the same kind of index for instantaneous coupling , and showed , remarkably , that the mutual information rate between @xmath0 and @xmath1 decomposes as the sum of the indices of causality from @xmath0 to @xmath1 and from @xmath1 to @xmath0 with the index of instantaneous coupling .
this decomposition was shown in the gaussian case , and it remains valid in any case when the indices of causality are replaced by transfer entropy rates , and the instantaneous coupling index is replaced by an instantaneous information exchange rate .
this link between granger causality and directed information theory was further supported by @xcite ( without mention of instantaneous coupling in @xcite ) , and the generalization to the nongaussian case by @xcite ( see also @xcite for related results ) .
however , _
prior _ to these recent studies , the generalization of geweke s idea to some general setting was reported in 1987 , in econometry by gouriroux _
@xcite , and in engineering by rissannen&wax @xcite .
gouriroux and his co - workers considered a joint markovian representation of the signals , and worked in a decision - theoretic framework .
they defined a sequence of nested hypotheses , whether causality was true or not , instantaneous coupling was present or not .
they then worked out the decision statistics using the kullback approach to decision theory @xcite , in which discrepancies between hypotheses are measured according to the kullback divergence between the probability measures under the hypotheses involved . in this setting , the decomposition obtained by geweke in the gaussian case was evidently generalised . in @xcite ,
the approach taken was closer to geweke s study , and it relied on system identification , in which the complexity of the model was taken into account .
the probability measures were parameterized , and an information measure that jointly assessed the estimation procedure and the complexity of the model was used when predicting a signal .
this allowed geweke s result to be extended to nonlinear modeling ( and hence the nongaussian case ) , and provided an information - theoretic interpretation of the tests .
once again , the same kind of decomposition of dependence was obtained by these authors .
we will see in section [ infotheory : sec ] that the decomposition holds due to kramers causal conditioning .
these studies were limited to the bivariate case @xcite . in the late 1990 s , some studies began to develop in the physics community on influences between dynamical systems . a first route was taken that followed the ideas of dynamic system studies for the prediction of chaotic systems . to determine if one signal influenced another , the idea was to consider each of the signals as measured states of two different dynamic systems , and then to study the master - slave relationships between these two systems ( for examples , see @xcite ) .
the dynamics of the systems was built using phase space reconstruction @xcite .
the influence of one system on another was then defined by making a prediction of the dynamics in the reconstructed phase space of one of the processes . to our knowledge , the setting was restricted to the bivariate case .
a second route , which was also restricted to the bivariate case , was taken and relied on information - theoretic tools .
the main contributions were from palu and schreiber @xcite , with further developments appearing some years later @xcite . in these studies ,
the influence of one process on the other was measured by the discrepancy between the probability measures under the hypotheses of influence or no influence . naturally , the measures defined very much resembled the measures proposed by gouriroux _ et
@xcite , and used the concept of conditional mutual information . the measure to assess whether one signal influences the other was termed _
transfer entropy _ by schreiber .
its definition was proposed under a markovian assumption , as was exactly done in @xcite .
the presentation by palu @xcite was more direct and was not based on a decision - theoretic idea .
the measure defined is , however , equivalent to the transfer entropy .
interestingly , palu noted in this 2001 paper the closeness of the approach to granger causality , as per the quotation : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` the [ latter ] measure can also be understood as an information theoretic formulation of the granger causality concept . ''
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ note that most of these studies considered bivariate analysis , with the notable exception of @xcite , in which the presence of side information ( other measured time series ) was explicitely considered . in parallel with these studies ,
many others were dedicated to the implementation of granger causality testing in fields as diverse as climatology ( with applications to the controversial questions of global warming ) and neuroscience ; see @xcite , to cite but a few . in a very different field ,
information theory , the problem of feedback has lead to many questions since the 1950 s .
we will not review or cite anything on the problem created by feedback in information theory as this is not within the scope of the present study , but some information can be found in @xcite .
instead , we will concentrate on studies that are directly related to the subject of this review .
a major breakthrough was achieved by james massey in 1990 in a short conference paper @xcite . following the ( lost ? ) ideas of marko on bidirectional information theory that were developed in the markovian case @xcite , massey re - examined the usual definition of what is called a discrete memoryless channel in information theory , and he showed that the usual definition based on some probabilistic assumptions prohibited the use of feedback .
he then clarified the definition of memory and feedback in a communication channel .
as a consequence , he showed that in a general channel used with feedback , the usual definition of capacity that relies on mutual information was not adequate .
instead , the right measure was shown to be _ directed information _ , an asymmetrical measure of the flow of information .
these ideas were further examined by kramer , who introduced the concept of causal conditioning , and who developed the first applications of directed information theory to communication in networks @xcite .
after some years , the importance of causal conditioning for the analysis of communication in systems with feedback was realized .
many studies were then dedicated to the analysis of the capacity of channels with feedback and the dual problem of rate - distortion theory @xcite . due to the rapid development in the study of networks ( _ e.g. , _ social networks , neural networks ) and of the afferent connectivity problem ,
more recently many authors made connections between information theory and granger causality @xcite .
some of these studies were restricted to the gaussian case , and to the bivariate case .
most of these studies did not tackle the problem of instantaneous coupling .
furthermore , several authors realized the importance of directed information theory to assess the circulation of information in networks @xcite .
tools from directed information theory appear as natural measures to assess granger causality . although granger causality can be considered as a powerful theoretical framework to study influences between signals mathematically , directed information theory provides the measures to test theoretical assertions practically . as already mentioned , these measures are transfer entropy ( and its conditional versions ) , which assesses the dynamical part of granger causality , and instantaneous information exchange ( and its conditional versions ) , which assesses instantaneous coupling .
this review is structured here as follows .
we will first give an overview of the definitions of granger causality .
these are presented in a multivariate setting .
we go gradually from weak definitions based on prediction , to strong definitions based on conditional independence .
the problem of instantaneous coupling is then discussed , and we show that there are two possible definitions for it . causality graphs ( after eichler @xcite ) provide particular reasons to prefer one of these definitions .
section [ infotheory : sec ] introduces an analysis granger causality from an information - theoretic perspective .
we insist on the concept of causal conditioning , which is at the root of the relationship studied .
section [ links : sec ] then highlights the links . here
, we first restate the definitions of granger causality using concepts from directed information theory . then from of a different point of view , we show how conceptual inference approaches lead to the measures defined in directed information theory .
the review then closes with a discussion of some of the aspects that we do not present here intentionally , and on some lines for further research .
all of the random variables , vectors and signals considered here are defined in a common probability space @xmath7 .
they take values either in @xmath8 or @xmath9 , @xmath10 being some strictly positive integer , or they can even take discrete values . as we concentrate on conceptual aspects rather than technical aspects , we assume that the variables considered are well behaved. in particular , we assume finiteness of moments of sufficient order .
we assume that continuously valued variables have a measure that is absolutely continuous with respect to the lebesgue measure of the space considered .
hence , the existence of probability density functions is assumed .
limits are supposed to exist when needed .
all of the processes considered in this report are assumed to be stationary .
we work with discrete time .
a signal will generically be denoted as @xmath11 .
this notation stands also for the value of the signal at time @xmath12 . the collection of successive samples of the signal
, @xmath13 will be denoted as @xmath14 . often , an initial time will be assumed .
this can be 0 , 1 , or @xmath15 . in any case
, if we collect all of the sample of the signals from the initial time up to time @xmath16 , we will suppress the lower index and write this collection as @xmath17 .
when dealing with multivariate signals , we use a graph - theoretic notation .
this will simplify some connections with graphical modeling .
let @xmath18 be an index set of finite cardinality @xmath19 .
@xmath20 is a @xmath10-dimensional discrete time stationary multivariate process for the probability space considered . for @xmath21
, @xmath22 is the corresponding component of @xmath23 .
likewise , for any subset @xmath24 , @xmath25 is the corresponding multivariate process @xmath26 .
we say that subsets @xmath27 form a partition of @xmath18 if they are disjoint and if @xmath28 . the information obtained by observing @xmath25 up to time @xmath12
is resumed by the filtration generated by @xmath29 .
this is denoted as @xmath30 .
furthermore , we will often identify @xmath25 with @xmath31 in the discussion .
the probability density functions ( p.d.f . ) or probability mass functions ( p.m.f ) will be denoted by the same notation as @xmath32 .
the conditional p.d.f . and p.m.f .
are written as @xmath33 .
the expected value is denoted as @xmath34,e_x[.]$ ] or @xmath35 $ ] if we want to specify which variable is averaged , or under which probability measure the expected value is evaluated .
independence between random variables and vectors @xmath0 and @xmath1 will be denoted as @xmath36 , while conditional independence given @xmath37 will be written as @xmath38 .
the early definitions followed the ideas of wiener : a signal @xmath0 causes a signal @xmath1 if the past of @xmath0 helps in the prediction of @xmath1 .
implementing this idea requires the performing of the prediction and the quantification of its quality .
this leads to a weak , but operational , form of the definitions of granger causality .
the idea of improving a prediction is generalized by encoding it into conditional dependence or independence .
consider a cost function @xmath39 ( @xmath12 is some appropriate dimension ) , and the associated risk @xmath40 $ ] , where @xmath41 stands for an error term .
let a predictor of @xmath42 be defined formally as @xmath43 , where @xmath31 and @xmath44 are subsets of @xmath18 , and @xmath5 is a function between appropriate spaces , chosen to minimize the risk with @xmath45 .
solvability may be granted if @xmath5 is restricted to an element of a given class of functions , such as the set of linear functions .
let @xmath46 be such a function class .
define : @xmath47 \label{risk : eq}\end{aligned}\ ] ] @xmath48 is therefore the optimal risk when making a one - step - ahead prediction of the multivariate signal @xmath49 from the past samples of the multivariate signal @xmath25 .
we are now ready to measure the influence of the past of a process on the prediction of another .
to be relatively general and to prepare comments on the structure of the graph , this can be done for subsets of @xmath18 .
we thus choose @xmath31 and @xmath44 to be two disjoint subsets of @xmath18 , and we define @xmath50 ( we use @xmath51 to mean substraction of a set ) .
we study causality from @xmath25 to @xmath49 by measuring the decrease in the quality of the prediction of @xmath42 when excluding the past of @xmath25 .
let @xmath52 be the optimal risk obtained for the prediction of @xmath49 from the past of all of the signals grouped in @xmath23 .
this risk is compared to @xmath53 , where the past of @xmath25 is omitted .
then , for the usual costs functions , we have necessarily : @xmath54 a natural first definition for granger causality is : @xmath25 granger does not cause @xmath49 relative to @xmath18 if and only if @xmath55 this definition of granger causality depends on the cost @xmath56 chosen as well as on the class @xmath46 of the functions considered . usually , a quadratic cost function is chosen , for its simplicity and for its evident physical interpretation ( a measure of the power of the error ) .
the choice of the class of functions @xmath46 is crucial .
the result of the causality test in definition 1 can change when the class is changed .
consider the very simple example of @xmath57 , where @xmath58 and @xmath59 are gaussian independent and identically distributed ( i.i.d . )
sequences that are independent of each other .
the covariance between @xmath60 and @xmath58 is zero , and using the quadratic loss and the class of linear functions , we conclude that @xmath1 does not granger cause @xmath0 , because using a linear function of @xmath61 to predict @xmath0 would lead to the same minimal risk as using a linear function of @xmath62 only .
however , @xmath58 obviously causes @xmath62 , but in a nonlinear setting .
the definition is given using the negative of the proposition . if by using the positive way ,
@xmath63 , granger proposes to say that @xmath25 is a _ prima facie _ cause of @xmath49 relative to @xmath18 , _ prima facie _ can be translated as at a first glance. this is used to insist that if @xmath18 is enlarged by including other measurements , then the conclusion might be changed .
this can be seen as redundant with the mention of the relativity to the observation set @xmath18 , and we therefore do not use this terminology . however , a mention of the relativity to @xmath18 must be used , as modification of this set can alter the conclusion .
a very simple example of this situation is the chain @xmath64 , where , for example , @xmath62 is an i.i.d .
sequence , @xmath65 , @xmath66 , @xmath67 being independent i.i.d . sequences .
relative to @xmath68 , @xmath0 causes @xmath37 if we use the quadratic loss and linear functions of the past samples of @xmath0 ( note here that the predictor @xmath69 must be a function of not only @xmath62 , but also of @xmath70 ) .
however , if we include the past samples of @xmath1 and @xmath71 , then the quality of the prediction of @xmath37 does not deteriorate if we do not use past samples of @xmath0 .
therefore , @xmath0 does not cause @xmath37 relative to @xmath71 .
the advantage of the prediction - based definition is that is leads to operational tests .
if the quadratic loss is chosen , working in a parameterized class of functions , such as linear filters or volterra filters , or even working in reproducing kernel hilbert spaces , allows the implementation of the definition @xcite . in such cases ,
the test needed can be evaluated efficiently from the data . from a theoretical point of view
, the quadratic loss can be used to find the optimal function in a much wider class of functions : the measurable functions . in this class , the optimal function for the quadratic loss is widely known to be the conditional expectation @xcite . when predicting @xmath49 from the whole observation set @xmath18 , the optimal predictor is written as @xmath72 $ ] .
likewise , elimination of @xmath31 from @xmath18 to study its influence on @xmath44 leads to the predictor @xmath73 $ ] , where @xmath74 .
these estimators are of little use , because they are too difficult , or even impossible , to compute .
however , they highlight the important of conditional distributions @xmath75 and @xmath76 in the problem of testing whether @xmath25 granger causes @xmath49 relative to @xmath18 or not .
the optimal predictors studied above are equal if the conditional probability distributions @xmath75 and @xmath76 are equal .
these distributions are identical if and only if @xmath77 and @xmath78 are independent conditionally to @xmath79 .
a natural extension of definition 1 relies on the use of conditional independence .
once again , let @xmath80 be a partition of @xmath18 .
@xmath25 does not granger cause @xmath49 relative to @xmath18 if and only if @xmath81 this definition means that conditionally to the past of @xmath82 , the past of @xmath25 does not bring more information about @xmath77 than is contained in the past of @xmath49 .
definition 2 is far more general than definition 1 .
if @xmath25 does not granger cause @xmath49 relatively to @xmath18 in the sense of definition 1 , it also does not in the sense of definition 2 .
then , definition 2 does not rely on any function class and on any cost function .
however , it lacks an inherent operational character : the tools to evaluate conditional independence remain to be defined .
the assessment of conditional independence can be achieved using measures of conditional independence , and some of these measures will be the cornerstone to link directed information theory and granger causality . note also that the concept of causality in this definition is again a relative concept , and that adding or deleting data from the observation set @xmath18 might modify the conclusions .
the definitions given so far concern the influence of the past of one process on the present of another one .
this is one reason that justifies the use of the term causality , when the definitions are actually based on statistical dependence . for an extensive discussion on the differences between causality and statistical dependence
, we refer to @xcite .
there is another influence between the processes that is not taken into account by definitions 1 and 2 .
this influence is referred to as instantaneous causality @xcite .
however , we will use our preferred term of instantaneous coupling , specifically to insist that it is not equivalent to a causal link _ per se _ , but actually a statistical dependence relationship . the term contemporaneous conditional independence that is used in @xcite could also be chosen .
instantaneous coupling measures the common information between @xmath83 and @xmath77 that is not shared with their past . a definition of instantaneous coupling
might then be that @xmath83 and @xmath77 are not instantaneously coupled if @xmath84 .
this definition makes perfect sense if the observation set is reduced to @xmath31 and @xmath44 , a situation we refer to as the bivariate case .
however , in general , there is also side information @xmath85 , and the definition must include this knowledge .
however , this presence of side information then leads to two possible definitions of instantaneous coupling .
@xmath25 and @xmath49 are not conditionally instantaneously coupled relative to @xmath18 if and only if @xmath86 , where @xmath87 is a partition of @xmath18 .
the second possibility is the following : @xmath25 and @xmath49 are not instantaneously coupled relative to @xmath18 if and only if @xmath88 note that definitions 3 and 4 are symmetrical in @xmath31 and @xmath44 ( the application of bayes theorem ) .
the difference between definitions 3 and 4 resides in the conditioning on @xmath89 instead of @xmath90 .
if the side information up to time @xmath16 is considered only as in definition 4 , the instantaneous dependence or independence is not conditional on the presence of the remaining nodes in @xmath85 .
thus , this coupling is a bivariate instantaneous coupling : it does measure instantaneous dependence ( or independence between @xmath31 and @xmath44 ) without considering the possible instantaneous coupling between either @xmath31 and @xmath85 or @xmath91 and @xmath85 .
thus , instantaneous coupling found with definition 4 between @xmath31 and @xmath44 does not preclude the possibility that the coupling is actually due to couplings between @xmath31 and @xmath85 and/or @xmath44 and @xmath85 .
inclusion of all of the information up to time @xmath92 in the conditioning variables allows the dependence or independence to be tested between @xmath83 and @xmath77 _ conditionally _ to @xmath93 .
we end up here with the same differences as those between correlation and partial correlation , or dependence and conditional independence for random variables . in graphical modeling , the usual graphs are based on conditional independence between variables @xcite .
these conditional independence graphs are preferred to independence graphs because of their geometrical properties ( _ e.g. , _ d - separation , @xcite ) , which match the markov properties possibly present in the multivariate distribution they represent . from a physical point of view , conditional independence might be preferable , specifically to eliminate false coupling due to third parties . in this respect ,
conditional independence is not the panacea , as independent variables can be conditionally dependent .
the well - known example is the conditional coupling of independent @xmath0 and @xmath1 by their addition . indeed , even if independent , @xmath0 and @xmath1 are conditionally dependent to @xmath94 .
granger causality graphs were defined and studied in @xcite .
a causality graph is a mixed graph @xmath95 that encodes granger causality relationships between the components of @xmath23 .
the vertex set @xmath18 stores the indexes of the components of @xmath23 .
@xmath96 is a set of directed edges beween vertices .
a directed edge from @xmath97 to @xmath98 is equivalent to `` @xmath22 granger causes @xmath99 relatively to @xmath18 '' .
@xmath100 is a set of undirected edges .
an undirected edge between @xmath22 and @xmath99 is equivalent to `` @xmath22 and @xmath99 are ( conditionally if def.4 adopted ) instantaneously coupled '' .
interestingly , a granger causality graph may have markov properties ( as in usual graphical models ) reflecting a particular ( spatial ) structure of the joint probability distribution of the whole process @xmath101 @xcite .
a taxonomy of markov properties : local , global , block recursive is studied in @xcite , and equivalence between these properties is put forward .
more interestingly , these properties are linked with topological properties of the graph . therefore ,
structural properties of the graphs are equivalent to a particular factorization of the joint probability of the multivariate process
. we will not continue on this subject here , but this must be known since it paves the way to more efficient inference methods for granger graphical modeling of multivariate processes ( see a first step in this direction in @xcite ) .
directed information theory is a recent extension of information theory , even if its roots go back to the 1960 s and 1970 s and the studies of marko @xcite .
the developments began in the late 1990 s , after the _ impetus _ given by james massey in 1990 @xcite .
the basic theory was then extended by gerhard kramer @xcite , and then further developed by many authors @xcite to cite a few .
we provide here a short review of the essentials of directed information theory .
we will , moreover , adopt a presentation close to the spirit of granger causality to highlight the links between granger causality and information theory .
we begin by recalling some basics from information theory .
then , we describe the information - theoretic approach to study directional dependence between stochastic processes , first in the bivariate case , and then , from section [ sideinfo : ssec ] , for networks , _
i.e. , _ the multivariate case . let @xmath102 $ ] be the entropy of a random vector @xmath78 , the density of which is @xmath103 .
let the conditional entropy be defined as @xmath104 $ ] .
the mutual information @xmath105 between @xmath78 and @xmath106 is defined as @xcite : @xmath107 where @xmath108 $ ] is the kulback - leibler divergence .
@xmath109 is 0 if and only if @xmath110 , and it is positive otherwise .
the mutual information effectively measures independence since it is 0 if and only if @xmath78 and @xmath106 are independent random vectors . as @xmath111
, mutual information can not handle directional dependence .
let @xmath112 be a third time series .
it might be a multivariate process that accounts for side information ( all of the available observations , but @xmath78 and @xmath106 ) . to account for @xmath112 ,
the conditional mutual information is introduced : @xmath113 \\ & = & d_{kl}\big(p(x_a^n , y_b^n , x_c^n ) ||
p(x_a^n | x_c^n ) p(y_b^n| x_c^n ) p(x_c^n ) \big)\end{aligned}\ ] ] @xmath114 is zero if and only if @xmath115 and @xmath116 are independent _ conditionally _ to @xmath112 .
stated differently , conditional mutual information measures the divergence between the actual observations and those which would be observed under the markov assumption @xmath117 .
arrows can be misleading here , as by reversibility of markov chains , the equality above holds also for @xmath118 .
this emphasizes how mutual information can not provide answers to the information flow directivity problem .
the dependence between the components of the stochastic process @xmath23 is encoded in the full generality by the joint probability distributions @xmath119 .
if @xmath18 is partitioned into subsets @xmath27 , studying dependencies between @xmath31 and @xmath44 then requires that @xmath119 is factorized into terms where @xmath25 and @xmath49 appear .
for example , as @xmath120 , we can factorize the probability distribution as @xmath121 , which appears to emphasize a link from @xmath31 to @xmath44 .
two problems appear , however : first , the presence of @xmath85 perturbs the analysis ( more than this , @xmath31 and @xmath85 have a symmetrical role here ) ; secondly , the factorization does not take into account the arrow of time , as the conditioning is considered over the whole observations up to time @xmath16 .
marginalizing @xmath82 out makes it possible to work directly on @xmath122 .
however , this eliminates all of the dependence between @xmath31 and @xmath44 that might exist _ via _
@xmath85 , and therefore this might lead to an incorrect assessment of the dependence . as for granger causality , this means that dependence analysis is relative to the observation set .
restricting the study to @xmath31 and @xmath44 is what we referred to as the bivariate case , and this allows the basic ideas to be studied .
we will therefore present directed information first in the bivariate case , and then turn to the full multivariate case .
the second problem is at the root of the measure of directional dependence between stochastic processes . assuming that @xmath123 and @xmath42 are linked by some physical ( _ e.g. , _ biological , economical ) system , it is natural to postulate that their dependence is constrained by causality : if @xmath124 , then an event occurring at some time in @xmath31 will influence @xmath44 later on
let us come back to the simple factorization above for the bivariate case .
we have @xmath125 , and furthermore : @xmath126 where for @xmath127 , the first term is @xmath128 .
the conditional distribution quantifies a directional dependence from @xmath31 to @xmath44 , but it lacks the causality property mentioned above , as @xmath129 quantifies the influence of the whole observation @xmath78 ( past and future of @xmath130 ) on the present @xmath131 knowing its past @xmath132 .
the causality principle would require the restriction of the _ prior _ time @xmath130 to the past of @xmath31 only .
kramer defined causal conditioning precisely in this sense @xcite . modifying eq .
( [ distcond : eq ] ) accordingly , we end up we the definition of the causal conditional probability distribution : @xmath133 remarkably this provides an alternative factorization of the joint probability . as noted by massey @xcite , @xmath134 can then be factorized as stands for the delayed collections of samples of @xmath49 .
if the time origin is finite , 0 or 1 , the first element of the list @xmath135 should be understood as a wild card @xmath136 which does not influence the conditioning . ] : @xmath137 assuming that @xmath25 is the input of a system that creates @xmath49 , @xmath138 characterizes the feedback in the system : each of the factors controls the probability of the input @xmath25 at time @xmath130 conditionally to its past and to the past values of the output @xmath49 .
likewise , the term @xmath139 characterizes the direct ( or feedforward ) link in the system .
several interesting simple cases occur : * in the absence of feedback in the link from @xmath31 to @xmath44 , there is the following : @xmath140 or equivalently , in terms of entropies , @xmath141 and as a consequence : @xmath142 * likewise , if there is only a feedback term , then @xmath143 and then : @xmath144 * if the link is memoryless , _
i.e. , _ the output @xmath49 does not depend on the past , then : @xmath145 these results allow the question of whether @xmath25 influences @xmath49 to be addressed .
if it does , then the joint distribution has the factorization of eq .
( [ factorisation : eq ] ) .
however , if @xmath25 does not influence @xmath49 , then @xmath146 , and the factorization of the joint probability distribution simplifies to @xmath147 .
kullback divergence between the probability distributions for each case generalizes the definition of mutual information to the directional mutual information : @xmath148 this quantity measures the loss of information when it is incorrectly assumed that @xmath25 does not influence @xmath49 .
this was called _ directed information _ by massey @xcite .
expanding the kullback divergence allows different forms for the directed information to be obtained : @xmath149 where we define the ` causal conditional entropy ' : @xmath150 \\ & = & \sum_{i=1}^n h\big ( x_b(i ) \big| x_b^{i-1 } , x_a^i \big ) \end{aligned}\ ] ] note that causal conditioning might involve more than one process .
this leads to the defining of the causal conditional directed information as : @xmath151 the basic properties of the directed information were studied by massey and kramer @xcite , and some are recalled below . as a kullback divergence ,
the directed information is always positive or zero .
then , simple algebraic manipulation allows the decomposition to be obtained : @xmath152 eq .
( [ decompdi : eq ] ) is fundamental , as it shows how mutual information splits into the sum of a feedforward information flow @xmath153 and a feedback information flow @xmath154 . in the absence of feedback , @xmath155 and @xmath156 .
( [ decompdi : eq ] ) allows the conclusion that the mutual information is always greater than the directed information , as @xmath157 is always positive or zero ( as directed information ) .
it is zero if and only if : @xmath158 or equivalently : @xmath159 this situation corresponds to the absence of feedback in the link @xmath124 , whence the fundamental result that the directed information and the mutual information are equal if the channel is free of feedback .
this result implies that mutual information over - estimates the directed information between two processes in the presence of feedback .
this was thoroughly studied in @xcite , in a communication - theoretic framework . the decomposition of eq .
( [ decompdi : eq ] ) is surprising , as it shows that the mutual information is not the sum of the directed information flowing in both directions .
instead , the following decomposition holds : @xmath160 where : @xmath161 this demonstrates that @xmath162 is symmetrical , but is in general not equal to the mutual information , except if and only if @xmath163 . as the term in the sum is the mutual information between the present samples of the two processes conditioned on their joint past values , this measure is a measure of instantaneous dependence .
it is indeed symmetrical in @xmath31 and @xmath44 .
the term @xmath164 will thus be named the _ instantaneous information exchange _ between @xmath25 and @xmath49 , and will hereafter be denoted as @xmath165 . like directed information ,
conditional forms of the instantaneous information exchange can be defined , as for example : @xmath166 which quantifies an instantaneous information exchange between @xmath31 and @xmath44 causally conditionally to @xmath85 .
entropy and mutual information in general increase linearly with the length @xmath16 of the recorded time series .
shannon s information rate for stochastic processes compensates for the linear growth by considering @xmath167 ( if the limit exists ) , where @xmath168 denotes any information measure on the sample @xmath17 of length @xmath16 . for the important class of stationary processes ( see _ e.g. , _
@xcite ) , the entropy rate turns out to be the limit of the conditional entropy : @xmath169 kramer generalized this result for causal conditional entropies @xcite , thus defining the directed information rate for stationary processes as : @xmath170 this result holds also for the instantaneous information exchange rate . note
that the proof of the result relies on the positivity of the entropy for discrete valued stochastic processes .
for continously valued processes , for which the entropy can be negative , the proof is more involved and requires the methods developed in @xcite , and see also @xcite .
as introduced by schreiber in @xcite , _ transfer entropy _ evaluates the deviation of the observed data from a model , assuming the following joint markov property : @xmath171 this leads to the following definition : @xmath172\end{aligned}\ ] ] then @xmath173 if and only if eq .
( [ eq : jointmarkov ] ) is satisfied .
although in the original definition , the past of @xmath0 in the conditioning might begin at a different time @xmath174 , for practical reasons @xmath175 is considered .
actually , no _ a priori _ information is available about possible delays , and setting @xmath175 allows the transfer entropy to be compared with the directed information . by expressing the transfer entropy as a difference of conditional entropies
, we get : @xmath176 for @xmath177 and choosing 1 as the time origin , the identity @xmath178 leads to : @xmath179 for stationary processes , letting @xmath180 and provided the limits exist , for the rates , we obtain : @xmath181 transfer entropy is the part of the directed information that measures the influence of the past of @xmath25 on the present of @xmath49 .
however it does not take into account the possible instantaneous dependence of one time series on another , which is handled by directed information .
moreover , as defined by schreiber in @xcite , only @xmath182 is considered in @xmath183 , instead of its sum over @xmath130 in the directed information .
thus stationarity is implicitly assumed and the transfer entropy has the same meaning as a rate . a sum over delays was considered by palu as a means of reducing errors when estimating the measure @xcite . summing over @xmath16 in eq .
( [ eq : schreibert ] ) , the following decomposition of the directed information is obtained : @xmath184 eq .
( [ dirinfodecomp : eq ] ) establishes that the influence of one process on another can be decomposed into two terms that account for the past and for the instantaneous contributions .
moreover , this explains the presence of the term @xmath185 in the r.h.s .
( [ sums2di : eq ] ) : instantaneous information exchange is counted twice in the l.h.s .
terms @xmath162 , but only once in the mutual information @xmath186 .
this allows eq .
( [ sums2di : eq ] ) to be written in a slightly different form , as : @xmath187 which is very appealing , as it shows how dependence as measured by mutual information decomposes as the sum of the measures of directional dependences and the measure of instantaneous coupling . the preceding developments aimed at the proposing of definitions of the information flow between @xmath25 and @xmath49 ; however , whenever @xmath31 and @xmath44 are connected to other parts of the network , the flow of information between @xmath31 and @xmath44 might be mediated by other members of the network .
time series observed on nodes other than @xmath31 and @xmath44 are hereafter referred to as side information .
the available side information at time @xmath16 is denoted as @xmath112 , with @xmath27 forming a partition of @xmath18 .
then , depending on the type of conditioning ( usual or causal ) two approaches are possible .
usual conditioning considers directed information from @xmath31 to @xmath44 that is conditioned on the whole observation @xmath112 . however , this leads to the consideration of causal flows from @xmath31 to @xmath44 that possibly include a flow that goes from @xmath31 to @xmath44 _ via _ @xmath85 in the future !
thus , an alternate definition for conditioning is required .
this is given by the definition of eq .
( [ causaldi1:eq ] ) of the causal conditional directed information : @xmath188 does the causal conditional directed information decompose as the sum of a causal conditional transfer entropy and a causal conditional instantaneous information exchange , as it does in the bivariate case ? applying twice the chain rule for conditional mutual information , we obtain : @xmath189 in this equation , @xmath190 is termed the causal conditional transfer entropy. this measures the flow of information from @xmath31 to @xmath44 by taking into account a possible route _ via _ @xmath85 . if the flow of information from @xmath31 to @xmath44 is entirely relayed by @xmath85 , the causal conditional transfer entropy is zero . in this situation ,
the usual transfer entropy is not zero , indicating the existence of a flow from @xmath31 to @xmath44 .
conditioning on @xmath85 allows the examination of whether the route goes through @xmath85 . the term : @xmath191 is the causal conditional information exchange. this measures the conditional instantaneous coupling between @xmath31 and @xmath44 .
the term @xmath192 emphasizes the difference between the bivariate and the multivariate cases .
this extra term measures an instantaneous coupling and is defined by : @xmath193 an alternate decomposition to eq .
( [ decomp : eq ] ) is : @xmath194
which emphasizes that the extra term comes from : @xmath195 this demonstrates that the definition of the conditional transfer entropy requires conditioning on the past of @xmath85 . if not , the extra term appears and accounts for instantaneous information exchanges between @xmath85 and @xmath44 , due to the addition of the term @xmath196 in the conditioning .
this extra term highlights the difference between the two different natures of instantaneous coupling . the first term
, @xmath197 describes the intrinsic coupling in the sense that it does not depend on parties other than @xmath85 and @xmath44 . the second coupling term , @xmath198 is relative to the extrinsic coupling , as it measures the instantaneous coupling at time @xmath130 that is created by variables other than @xmath44 and @xmath85 .
as discussed in section [ instantcoupl : ssec ] , the second definition for instantaneous coupling considers conditioning on the past of the side information _
only_. causally conditioning on @xmath199 does not modify the results of the bivariate case . in particular , we still get the elegant decomposition : @xmath200 and therefore , the decomposition of eq .
( [ sums2diinst : eq ] ) is generalized to : @xmath201 where : @xmath202 is the causally conditioned mutual information . finally , let us consider that for jointly stationary times series , the causal directed information rate is defined similarly to the bivariate case , as : @xmath203 in this section we have emphasized on kramer s causal conditioning , both for the definition of directed information and for taking into account side information .
we have also shown that schreiber s transfer entropy is that part of the directed information that is dedicated to the strict sense of causal information flow ( not accounting for simultaneous coupling ) .
the next section more explicitely revisits the links between granger causality and directed information theory .
granger causality in its probabilistic form is not operational . in practical situations , for assessing granger causality between time series
, we can not use the definition directly .
we have to define dedicated tools to assess the conditional independence .
we use this inference framework to show the links between information theory and granger causality .
we begin by re - expressing granger causality definitions in terms of some measures that arise from directed information theory .
therefore , in an inference problem , these measures can be used as tools for inference .
however , we show in the following sections that these measures naturally emerge from the more usual statistical inference strategies . in the following , and as above
, we use the same partitioning of @xmath18 into the union of disjoint subsets of @xmath31 , @xmath44 and @xmath85 .
as anticipated in the presentation of directed information , there are profound links between granger causality and directed information measures .
granger causality relies on conditional independence , and it can also be defined using measures of conditional independence .
information - theoretic measures appear as natural candidates .
recall that two random elements are independent if and only if their mutual information is zero .
moreover , two random elements are independent conditionally to a third one if and only if the conditional mutual information is zero
. we can reconsider definitions 2 , 3 and 4 and recast them in term of information - theoretic measures .
definition 2 stated that @xmath25 does not granger cause @xmath49 relative to @xmath18 if and only if @xmath204 .
this can be alternatively rephrased into : @xmath25 does not granger cause @xmath49 relative to @xmath18 if and only if @xmath205 since @xmath206 is equivalent to @xmath207 .
otherwise stated , the transfer entropy from @xmath31 to @xmath44 causally conditioned on @xmath85 is zero if and only if @xmath31 does not granger cause @xmath44 relative to @xmath18 .
this shows that causal conditional transfer entropy can be used to assess granger causality .
likewise , we can give alternative definitions of instantaneous coupling .
@xmath25 and @xmath49 are not conditionally instantaneously coupled relative to @xmath18 if and only if @xmath208 , or if and only if the instantaneous information exchange causally conditioned on @xmath85 is zero .
the second possible definition of instantaneous coupling is equivalent to : @xmath25 and @xmath49 are not instantaneously coupled relative to @xmath18 if and only if @xmath209 , or if and only if the instantaneous information exchange causally conditioned on the past of @xmath85 is zero .
note that in the bivariate case only ( when @xmath85 is not taken into account ) , the directed information @xmath210 summarizes both the granger causality and the coupling , as it decomposes as the sum of the transfer entropy @xmath211 and the instantaneous information exchange @xmath212 .
we consider the practical problem of inferring the graph of dependence between the components of a multivariate process .
let us assume that we have measured a multivariate process @xmath213 for @xmath214 .
we want to study the dependence between each pair of components ( granger causality and instantaneous coupling between any pair of components relative to @xmath18 ) .
we can use the result of the preceding section to evaluate the directed information measures on the data .
when studying the influence from any subset @xmath31 to any subset @xmath44 , if the measures are zero , then there is no causality ( or no coupling ) ; if they are strictly positive , then @xmath31 granger causes @xmath44 relative to @xmath18 ( or @xmath31 and @xmath44 are coupled relative to @xmath18 ) .
this point of view has been adopted in many of the studies that we have already referred to ( _ e.g. _ @xcite ) , and it relies on estimating the measures from the data .
we will not review the estimation problem here .
however , it is interesting to examine more traditional frameworks for testing granger causality , and to examine how directed information theory naturally emerges from these frameworks . to begin with
, we show how the measures defined emerge from a binary hypothesis - testing view of granger causality inference .
we then turn to prediction and model - based approaches .
we will review how geweke s measures of granger causality in the gaussian case are equivalent to directed information measures .
we will then present a more general case adopted by @xcite and based on a model of the data .
in the inference problem , we want to determine whether or not @xmath25 granger causes ( is coupled with ) or not @xmath49 relative to @xmath18 .
this can be formulated as a binary hypothesis testing problem . for inferring dependencies between @xmath31 and @xmath44 relative to @xmath18 , we can state the problem as follows .
assume we observe @xmath215 .
then , we want to test : @xmath25 does not granger cause @xmath49 , against @xmath25 causes @xmath49 ; and @xmath25 and @xmath49 are instantaneously coupled against ` @xmath25 are @xmath49 not instantaneously coupled ' .
we will refer to the first test as the granger causality test , and to the second one , as the instantaneous coupling test .
in the bivariate case , for which the granger causality test indicates : @xmath216 this leads to the testing of different functional forms of the conditional densities of @xmath131 given the past of @xmath25 .
the likelihood of the observation under @xmath217 is the full joint probability @xmath218 .
under @xmath219 we have @xmath220 and the likelihood reduces to @xmath221 . the log likelihood
_ ratio _ for the test is : @xmath222 for example , in the case where the multivariate process is a positive harris recurrent markov chain @xcite , the law of large numbers applies and we have under hypothesis @xmath217 : @xmath223 where @xmath224 is the transfer entropy rate .
thus from a practical point of view , as the amount of data increases , we expect the log likelihood _
ratio _ to be close to the transfer entropy rate ( under @xmath217 ) .
turning the point of view , this can justify the use of an estimated transfer entropy to assess granger causality . under @xmath219
, @xmath225 converges to @xmath226 , which can be termed ` the lautum transfer entropy rate ' that extends the ` lautum directed information ' defined in @xcite .
directed information can be viewed as a measure of the loss of information when assuming @xmath25 does not causally influence @xmath49 when it actually does . likewise , ` lautum directed information ' measures the loss of information when assuming @xmath25 does causally influence @xmath49 , when actually it does not . for testing instantaneous coupling , we will use the following : @xmath227 where under @xmath219 , there is no coupling .
then , under @xmath217 and some hypothesis on the data , the likelihood ratio converges almost surely to the information exchange rate @xmath228
. a related encouraging result due to @xcite is the emergence of the directed information in the false - alarm probability error rate . merging the two tests ( [ test1:eq]),([test2:eq ] ) , _
i.e. , _ testing both for causality and coupling , or neither , the test is written as : @xmath229 among the tests with a probability of miss @xmath230 that is lower than some positive value @xmath231 , the best probability of false alarm @xmath232 follows @xmath233 when @xmath183 is large .
for the case studied here , this is the so - called stein lemma @xcite . in the multivariate case , there is no such result in the literature .
an extension is proposed here .
however , this is restricted to the case of instantaneously _ uncoupled _ time series .
thus , we assume for the end of this subsection that : @xmath234 which means that there is no instantaneous exchange of information between the three subsets that form a partition of @xmath18 . this assumption has held in most of the recent studies that have applied granger causality tests
. it is , however , unrealistic in applications where the dynamics of the processes involved are faster than the sampling period adopted ( see @xcite for a discussion in econometry ) .
consider now the problem of testing granger causality of @xmath31 on @xmath44 relative to @xmath18 .
the binary hypothesis test is given by : @xmath235 the log likelihood _
ratio _ reads as : @xmath236 again , by assuming that the law of large numbers applies , we can conclude that under @xmath217 @xmath237 this means that the causal conditional transfer entropy rate is the limit of the log likelihood _
ratio _ as the amount of data increases .
following definition 1 and focusing on the quadratic risk @xmath238 $ ] , geweke introduced the following indices for the study of gaussian stationary processes @xcite :
@xmath239 geweke demonstrated the efficiency of these indices for testing granger causality and instantaneous coupling ( bivariate and multivariate cases ) .
furthermore , in the bivariate case , geweke showed that : @xmath240 where @xmath241 is the mutual information rate .
this relationship that was already sketched out in @xcite , is nothing but eq .
( [ sums2diinst : eq ] ) .
indeed , in the gaussian case , @xmath242 and @xmath243 stem from the knowledge that the entropy rate of a gaussian stationary process is the logarithm of the asymptotic power of the one - step - ahead prediction @xcite .
likewise , we can show that @xmath244 and @xmath245 holds .
in the multivariate case , conditioning on the past of the side information , _
i.e. _ @xmath199 , in the definition of @xmath246 , a decomposition analagous to eq .
( [ gewedecomp : eq ] ) holds , and is exactly that given by eq .
( [ sums2diinstcond : eq ] ) . in a more general framework
, we examine how a model - based approach can be used to test for granger causality , and how directed information comes into play .
let us consider a rather general model in which @xmath247 is a multivariate markovian process that statisfies : @xmath248 where @xmath249 is a function belonging to some functional class @xmath46 , and where @xmath250 is a multivariate i.i.d .
sequence , the components of which are not necessarily mutually independent .
function @xmath251 might ( or might not ) dependon @xmath252 , a multidimensional parameter .
this general model includes as a particular case , linear multivariate autoregressive with moving average ( arma ) models , and nonlinear arma models ; @xmath251 can also stand for a function belonging to some reproducing kernel hilbert space , which can be estimated from the data @xcite . using the partition @xmath27 ,
this model can be written equivalently as : @xmath253 where the functions @xmath254 are the corresponding components of @xmath255 .
this relation can be used for inference in a parametric setting : the functional form is assumed to be known and the determination of the function is replaced by the estimation of the parameters @xmath256 .
this can also be used in a nonparametric setting , in which case the function @xmath5 is searched for in an appropriate functional space , such as an rkhs associated to a kernel @xcite . in any case , for studying the influence of @xmath25 to @xmath49 relative to @xmath18 , two models are required for @xmath49 : one in which @xmath49 explicitly depends on @xmath25 , and the other one in which @xmath49 does not depend on @xmath25 . in the parametric
setting , the two models can be merged into a single model , in such a way that some components of the parameter @xmath257 are , or not , zero , which dependis whether @xmath31 causes @xmath44 or not .
the procedure then consists of testing nullity ( or not ) of these components . in the linear gaussian case ,
this leads to the geweke indices discussed above . in the nonlinear ( nongaussian ) case
, the geweke indices can be used to evaluate the prediction in some classes of nonlinear models ( in the minimum mean square error sense ) . in this latter case , the decomposition of the mutual information , eq .
( [ gewedecomp : eq ] ) , has no reason to remain valid .
another approach base relies on directly modeling the probability measures .
this approach has been used recently to model spiking neurons and to infer granger causality between several neurons working in the class of generalized linear models @xcite .
interestingly , the approach has been used either to estimate the directed information @xcite or to design a likelihood ratio test @xcite .
suppose we wish to test whether @xmath25 granger causes @xmath49 relative to @xmath18 as a binary hypothesis problem , as in section [ binaryhyp : ssec ] .
forgetting the problem of instantaneous coupling , the problem is then to choose between the hypotheses : @xmath258 where the existence of causality is entirely reflected into the parameter @xmath252 . to be more precise ,
@xmath259 should be seen as a restriction of @xmath260 when its components linked to @xmath25 are set to zero . as a simple example using the model approach discussed above , consider the simple linear gaussian model @xmath261 where @xmath262 is an i.i.d .
gaussian sequence , and @xmath263 are multivariate impulse responses of appropriate dimensions .
define @xmath264 and @xmath265 .
testing for granger causality is then equivalent to testing @xmath266 ; furthermore , the likelihood _ ratio _ can be implemented due to the gaussian assumption .
the example developed in @xcite , assumes that the probability that neuron @xmath98 ( @xmath267 ) sends a message at time @xmath268 ( @xmath269 ) to its connected neighbors is given by the conditional probability @xmath270 where @xmath271 is some decision function , the output of which belongs to @xmath272 $ ] , @xmath31 represents the subset of neurons that can send information to @xmath98 , and @xmath273 represents external inputs to @xmath98 . defining this probability for all @xmath274
completely specifies the behavior of the neural network @xmath18 .
the problem is a composite hypothesis testing problem , in which parameters defining the likelihoods have to be estimated .
it is known that tere is no definitive answer to this problem @xcite .
an approach that relies on an estimation of the parameters using maximum likelihood can be used .
letting @xmath275 be the space where parameter @xmath252 is searched for and @xmath276 the subspace where @xmath259 lives , then the generalized loglikelihood _ ratio _ test reads : @xmath277 where @xmath278 denotes the maximum likelihood estimator of @xmath252 under hypothesis @xmath130 .
in the linear gaussian case , we will recover exactly the measures developed by geweke . in a more general case , and as illustrated in section [ binaryhyp : ssec ] , as the the maximum likelihood estimates are efficient , we can conjecture that the generalized log likelihood _ ratio _ will converge to the causal conditional transfer entropy rate if sufficiently relevant conditions are imposed on the models ( _ e.g. , _ markov processes with recurrent properties ) .
this approach was described in @xcite in the bivariate case .
granger causality was developed originally in econometrics , and it is now transdisciplinary , with the literature on the subject is widely dispersed . we have tried here to sum up the profound links that exist between granger causality and directed information theory .
the key ingredients to build these links are conditional independence and the recently introduced causal conditioning .
we have eluded the important question of how to practically use the definitions and measures presented here . some of the measures can be used and implemented easily , especially in the linear gaussian case . in a more general case , different approaches can be taken .
the information - theoretic measures can be estimated , or the prediction can be explicitly carried out and the residuals used to assess causality .
many studies have been carried out over the last 20 years on the problem of estimation of information - theoretic measures .
we refer to @xcite for information on the different ways to estimate information measures .
recent studies into the estimation of entropy and/or information measures are @xcite .
the recent report by @xcite extensively details and applies transfer entropy in neuroscience using @xmath12-nearest neighbors type of estimators .
concerning the applications , important reviews include @xcite , where some of the ideas discussed here are also mentioned , and where practicalities such as the use of surrogate data , for example , are extensively discussed .
applications for neuroscience are discussed in @xcite .
information - theoretic measures of conditional independence based on kullback divergence were chosen here to illustrate the links between granger causality and ( usual ) directed information theory . other type of divergence could have been chosen ; metrics in probability space could also be useful in the assessing of conditional independence . as an illustration , we refer to the study of fukumizu and co - workers @xcite , where conditional independence was evaluated using the hilbert - schmidt norm of an operator between reproducing kernel hilbert spaces .
the operator generalizes the partial covariance between two random vectors given a third one , and is called the conditional covariance operator .
furthermore , the hilbert - schmidt norm of conditional covariance operator can be efficiently estimated from data . a related approach is also detailed in @xcite .
many important directions can be followed .
causality between nonstationary processes has rarely been considered ( see however @xcite for an _ ad - hoc _ approach in neuroscience ) .
a very promising methodology is to adopt a graphical modeling way of thinking .
the result of @xcite on the structural properties of markov - granger causality graphs can be used to identify such graphs from real datasets . a first step in this direction
was proposed by @xcite .
assuming that the network under study is a network of sparsely connected nodes and that some markov properties hold , efficient estimation procedures can be designed , as is the case in usual graphical modeling .
p.o.a . is supported by a marie curie international outgoing fellowship from the european community .
p. o. amblard and o. j. j. michel .
sur diffrentes mesures de dpendance causales entre signaux al atoires ( on different measures of causal dependencies between random signals ) . in _ proc .
gretsi , dijon , france , sept .
_ , 2009 .
m. kaminski , m. ding , w. truccolo , and s. bressler .
evaluating causal relations in neural systems : granger causality , directed transfer functions and statistical assessment of significance .
, 85:145157 , 2001 . c. j. quinn and t. p. coleman and n. kiyavash and n. g.hastopoulos .
estimating the directed information to infer causal relationships in ensemble neural spike train recordings , journal of computational neuroscience , 30 : 1744 , 2011 | this report reviews the conceptual and theoretical links between granger causality and directed information theory .
we begin with a short historical tour of granger causality , concentrating on its closeness to information theory .
the definitions of granger causality based on prediction are recalled , and the importance of the observation set is discussed .
we present the definitions based on conditional independence . the notion of instantaneous coupling
is included in the definitions .
the concept of granger causality graphs is discussed .
we present directed information theory from the perspective of studies of causal influences between stochastic processes .
causal conditioning appears to be the cornerstone for the relation between information theory and granger causality . in the bivariate case ,
the fundamental measure is the directed information , which decomposes as the sum of the transfer entropies and a term quantifying instantaneous coupling .
we show the decomposition of the mutual information into the sums of the transfer entropies and the instantaneous coupling measure , a relation known for the linear gaussian case .
we study the multivariate case , showing that the useful decomposition is blurred by instantaneous coupling .
the links are further developed by studying how measures based on directed information theory naturally emerge from granger causality inference frameworks as hypothesis testing . *
keyword : * granger causality , transfer entropy , information theory , causal conditioning , conditional independence |
the growing abundance of ( mobile ) computation and communication devices creates a rich potential for novel distributed systems and applications . unlike classical networks ,
often the resulting networks and applications are characterized by a high level of churn and , especially in the case of mobile devices , a potentially constantly changing topology .
traditionally , changes in a network have been studied as faults or as exceptional events that have to be tolerated and possibly repaired .
however , particularly in mobile applications , dynamic networks are a typical case and distributed algorithms have to properly work even under the assumption that the topology is constantly changing .
consequently , in the last few years , there has been an increasing interest in distributed algorithms that run in dynamic systems .
specifically , a number of recent papers investigate the complexity of solving fundamental distributed computations and information dissemination tasks in dynamic networks , e.g. , @xcite . particularly important in the context of this paper
is the synchronous , adversarial dynamic network model defined in @xcite .
while the network consists of a fixed set of participants @xmath16 , the topology can change arbitrarily from round to round , subject to the restriction that the network of each round needs to be connected or satisfy some stronger connectivity requirement .
we study lower bounds on the problem of disseminating a bunch of tokens ( messages ) to all the nodes in a dynamic network as defined in @xcite .
initially @xmath0 tokens are placed at some nodes in the network .
time is divided into synchronous rounds , the network graph of every round is connected , and in every round , each node can broadcast one token to all its neighbors . if in addition , all nodes know the size of the network @xmath1 , we can use the following basic protocol to broadcast all @xmath0 tokens to all the nodes .
the tokens are broadcast one after the other such that for each token during @xmath17 rounds , every node that knows about the token forwards it . because in each round , there has to be an edge between the nodes knowing the token and the nodes not knowing it , at least one new node receives the token in every round and thus , after @xmath17 rounds ,
all nodes know the token .
assuming that only one token can be broadcast in a single message , the algorithm requires @xmath18 rounds to disseminate all @xmath0 tokens to all the nodes . even though the described approach seems almost trivial , as long as we do not consider protocols based on network coding , @xmath2 is the best upper bound known . in @xcite ,
token - forwarding algorithm _ is defined as an algorithm that needs to forward tokens as they are and is not allowed to combine or change tokens in any way .
note that the algorithm above is a token - forwarding algorithm . in a recent paper , dutta et al .
show that for deterministic token - forwarding algorithms , the described simple strategy indeed can not be significantly improved by showing a lower bound of @xmath19 rounds @xcite .
their lower bound is based on the following observation .
assume that initially , every node receives every token for free with probability @xmath20 ( independently for all nodes and tokens ) .
now , with high probability , whatever tokens the nodes decide to broadcast in the next round , the adversary can always find a graph in which new tokens are learned across at most @xmath21 edges .
hence , in each round , at most @xmath21 tokens are learned . because also after randomly assigning tokens with probability @xmath20 , overall still roughly @xmath22 tokens are missing , the lower bound follows . we extend the lower bound from @xcite in various natural directions .
specifically , we make the contributions listed in the following .
all our lower bounds hold for deterministic algorithms and for randomized algorithms assuming a strongly adaptive adversary ( cf.section [ sec : model ] ) .
our results are also summarized in which is discussed in section [ sec : related ] .
[ [ multiple - tokens - per - round ] ] multiple tokens per round : + + + + + + + + + + + + + + + + + + + + + + + + + + assume that instead of forwarding a single token per round , each node is allowed to forward up to @xmath23 tokens in each round . in the simple token - forwarding algorithm that we described above
, we can then forward a block of @xmath24 tokens to every node in @xmath17 rounds and we therefore get an @xmath25 round upper bound . we show that every ( randomized ) token - forwarding algorithm needs at least @xmath26 rounds .
[ [ interval - connectivity ] ] interval connectivity : + + + + + + + + + + + + + + + + + + + + + + it is natural to assume that a dynamic network can not change arbitrarily from round to round and that some paths remain stable for a while .
this is formally captured by the notion of interval connectivity as defined in @xcite .
a network is called @xmath10-interval connected for an integer parameter @xmath27 if for any @xmath10 consecutive rounds , there is a stable connected subgraph .
it is shown in @xcite that in a @xmath10-interval connected dynamic network , @xmath0-token dissemination can be solved in @xmath28 rounds . in this paper , we show that every ( randomized ) token - forwarding algorithm needs at least @xmath29 rounds .
[ [ vertex - connectivity ] ] vertex connectivity : + + + + + + + + + + + + + + + + + + + + if instead of merely requiring that the network is connected in every round , we assume that the network is @xmath7-vertex connected in every round for some @xmath30 , we can also obtain a speed - up . because in a @xmath7-vertex connected graph , every vertex cut has size at least @xmath7 , if in a round all nodes that know a token @xmath31 broadcast it , at least @xmath7 new nodes are reached .
the basic token - forwarding algorithm thus leads to an @xmath32 upper bound .
we prove this upper bound tight up to a small factor by showing an @xmath33 lower bound .
[ [ delta - partial - token - dissemination ] ] @xmath13-partial token dissemination : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + finally we consider the basic model , but relax the requirement on the problem by requiring that every node needs to obtain only a @xmath13-fraction of all the @xmath0 tokens for some parameter @xmath34 $ ] .
we show that even then , at least @xmath35 rounds are needed .
this also has implications for algorithms that use forward error correcting codes ( fec ) to forward coded packets instead of tokens .
we show that such algorithms still need at least @xmath36 rounds until every node has received enough coded packets to decode all @xmath0 tokens .
.upper and lower bounds for token forwarding ( tf ) algorithms and network coding ( nc ) based solutions ( bounds in bold are proven in this paper ) .
all tf algorithms are distributed and deterministic while all lower bounds are for centralized randomized algorithms and a strongly adaptive adversary . the nc algorithms work either in the distributed setting against a ( standard ) adaptive adversary ( 1 ) or in the centralized setting against a strongly adaptive adversary ( 2 ) . [ cols="<,^,^,^",options="header " , ] + as stated in the introduction , we use the network model introduced in @xcite .
that paper studies the complexity of computing basic functions such as counting the number of nodes in the network , as well as the cost the token dissemination problem that we investigate in the present paper .
previously , some basic results of the same kind were also obtained in @xcite for a similar network model . the token dissemination problem as studied here
is first considered in @xcite in a dynamic network setting .
the paper gives a variant of the distributed @xmath2 token - forwarding algorithm for the case when the number of nodes @xmath1 is not known .
it is also shown that @xmath10-interval connectivity and always @xmath7-vertex connectivity are interesting parameters that speed up the solution by factors of @xmath37 and @xmath38 , respectively .
in addition , @xcite gives a first @xmath39 lower bound for token - forwarding algorithms in the centralized setting we study in the present paper . that lower bound is substantially improved in @xcite , where an almost tight @xmath19 lower bound is proven .
as the lower bound from @xcite is the basis of our results , we discuss it in detail in section [ sec : original ] . the fastest known algorithms for token dissemination in dynamic networks are based on random linear network coding .
there , tokens are understood as elements ( or vectors ) of a finite field and in every round , every node broadcasts a random linear combination of the information it possesses . in a centralized setting , the overhead for transmitting the coefficients of the linear combination can be neglected . for this case , it is shown in @xcite that in always connected dynamic networks , @xmath0 tokens can be disseminated in optimal @xmath40 time . if messages are large enough to store @xmath24 tokens , this bound improves to again optimal @xmath41 time .
it is also possible to extend these results to always @xmath7-connected networks and to the partial token dissemination problem .
note that one possible solution for @xmath13-partial token dissemination is to solve regular token dissemination for only @xmath42 tokens .
if the overhead for coefficients is not neglected , the best known upper bounds are given in @xcite .
the best bounds for tokens of size @xmath21 , as well as the upper and lower bounds for the other scenarios are listed in .
the given bound for always @xmath7-vertex connected networks is not proven in @xcite , it can however be obtained with similar techniques .
note also that instead of @xmath10-interval connectivity , @xcite considers a somewhat stronger assumption called @xmath10-stability . in a @xmath10-stable network ,
the network remains fixed for intervals of length @xmath10 .
apart from token dissemination and basic aggregation tasks , other problems have been considered in the same or similar adversarial dynamic network models . in @xcite , the problem of coordinating actions in time
is studied for always connected dynamic networks . in a recent paper ,
bounds on what can be achieved if the network is not always connected are discussed in @xcite . for a model where nodes know their neighbors before communicating , @xcite studies the time to do a random walk if the network can change adversarially .
further , the problem of gradient clock synchronization has been studied for an asynchronous variant of the model @xcite .
in addition , a number of papers investigate a radio network variant of essentially the dynamic network model studied here @xcite .
another line of research looks at random dynamic networks that result from some markov process , e.g. , @xcite .
mostly these papers analyze the time required to broadcast a single message in the network . for a more thorough discussion of related work
, we refer to a recent survey @xcite .
lower bounds : + @xmath43 : @xcite @xmath44 ( for very restricted class of algorithms ) : @xcite @xmath45 : @xcite all lower bounds hold for centralized algorithms upper bounds : @xmath44 @xcite c - connected : @xmath46 ( easy extension of @xcite ) @xmath13-gossip : @xmath47 ( easy extension of @xcite ) t - interval - connectivity and t - stability : @xmath48 @xcite + nc : @xmath49 ( packet size = token size = @xmath21 ) nc t - stability : @xmath50 @xcite nc token size / packet size = l : @xmath51 @xcite nc centralized : @xmath52 @xcite nc centralized c - connected : @xmath53 @xcite network coding @xcite , random linear network coding @xcite
in this section we introduce the dynamic network model and the token dissemination problem . [ [ dynamic - networks ] ] dynamic networks : + + + + + + + + + + + + + + + + + we follow the dynamic network model of @xcite : a dynamic network consists of a fixed set @xmath16 of @xmath1 nodes and a dynamic edge set @xmath54 .
time is divided into synchronous rounds so that the network graph of round @xmath55 is @xmath56 .
we use the common assumption that round @xmath57 starts at time @xmath58 and it ends at time @xmath57 . in each round @xmath57 , every node @xmath59 can send a message to all its neighbors in @xmath60 . note that we assume that @xmath61 has to send the same message to all neighbors , i.e. , communication is by local broadcast .
also , we assume that at the beginning of a round @xmath57 , when the messages are chosen , nodes are not aware of their neighborhood in @xmath60 .
we typically assume that the message size is bounded by the size of a fixed number of tokens .
we say that a dynamic network @xmath62 is _ always @xmath7-vertex connected _ iff @xmath60 is @xmath7-vertex connected for every round @xmath57 .
if a network @xmath63 is always @xmath64-vertex connected , we also say that @xmath63 is _
always connected_. further , we use the definition for interval connectivity from @xcite .
a dynamic network is @xmath10-interval connected for an integer parameter @xmath27 iff the graph @xmath65 is connected for every @xmath55 .
hence , a graph is @xmath10-interval connected iff there is a stable connected subgraph for every @xmath10 consecutive rounds .
note we do not assume that nodes know the stable subgraph . also note that a dynamic graph is @xmath64-interval connected iff it is always connected . for our lower bound ,
we assume randomized algorithms and a _ strongly adaptive adversary _ which can decide on the network @xmath60 of round @xmath57 based on the complete history of the network up to time @xmath58 as well as on the messages the nodes send in round @xmath57 .
note that the adversary is stronger than the more typical _ adaptive adversary _ where the graph @xmath60 of round @xmath57 is independent of the random choices that the nodes make in round @xmath57 .
[ [ the - token - dissemination - problem ] ] the token dissemination problem : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we prove lower bounds on the following _ token dissemination problem_. there are @xmath0 tokens initially distributed among the nodes in the network ( for simplicity , we assume that @xmath0 is at most polynomial in @xmath1 ) .
we consider _ token - forwarding algorithms _ as defined in @xcite . in each round , every node is allowed to broadcast @xmath66 of the tokens it knows to all neighbors . except for section [ sec : multipletokens ] ,
we assume that @xmath67 .
no other information about tokens can be sent , so that a node @xmath68 knows exactly the tokens @xmath68 kept initially and the tokens that were included in some message @xmath68 received .
in addition , we also consider the _
@xmath13-partial token dissemination _ problem .
again , there are @xmath0 tokens that are initially distributed among the nodes in the network . but here , the requirement is weaker and we only demand that in the end , every node knows a @xmath13-fraction of the @xmath0 tokens for some @xmath69 $ ] .
we prove our lower bounds for centralized algorithms where a central scheduler can determine the messages sent by each node in a round @xmath57 based on the initial state of all the nodes before round @xmath57 .
note that lower bounds obtained for such centralized algorithms are stronger than lower bounds for distributed protocols where the message broadcast by a node @xmath68 in round @xmath57 only depends on the initial state of @xmath68 before round @xmath57 .
we start our description of the lower bound by outlining the basic techniques and by giving a slightly polished version of the lower bound proof by dutta et al . @xcite . for the discussion here , we assume that in each round , each node is allowed to broadcast a single token , i.e. , @xmath67 . in the following ,
we make the standard assumption that round @xmath57 lasts from time @xmath58 to time @xmath57 . for each node , we maintain two sets of tokens . for a time @xmath70 and a node @xmath68 , let @xmath71 be the set of tokens known by node @xmath68 at time @xmath31 .
in addition the adversary determines a token set @xmath72 for every node , where @xmath73 for all @xmath70 .
the sets @xmath72 are constructed such that under the assumption that each node @xmath68 knows the tokens @xmath74 at time @xmath31 , in round @xmath75 , overall the nodes can not learn many new tokens .
specifically , we define a potential function @xmath76 as follows : @xmath77 note that for the token dissemination problem to be completed at time @xmath10 it is necessary that @xmath78 .
assume that at the beginning , the nodes know at most @xmath79 tokens on average , i.e. , @xmath80 .
for always connected dynamic graphs , we will show that there exists a way to choose the @xmath81-sets such that @xmath82 and that for every choice of the algorithm , a simple greedy adversary can ensure that the potential grows by at most @xmath21 per round .
we then have @xmath83 and since the potential needs to grow to @xmath44 , we get an @xmath84 lower bound . in each round @xmath57 , for each node @xmath68 , an algorithm can decide on a token to send .
we denote the token sent by node @xmath68 in round @xmath57 by @xmath85 and we call the collection of pairs @xmath86 for nodes @xmath87 , the token assignment of round @xmath57 .
note that because a node can only broadcast a token it knows , @xmath88 needs to hold .
however , for most of the analysis , we do not make use of this fact and just consider all the @xmath0 possible pairs @xmath86 for a node @xmath68 . if the graph @xmath60 of round @xmath57 contains the edge @xmath89 , @xmath68 or @xmath61 learns a new token if @xmath90 or if @xmath91 .
moreover , the edge @xmath89 contributes to an increase of the potential function @xmath92 in round @xmath57 if @xmath93 or if @xmath94 .
we call an edge @xmath95 _ free _ in round @xmath57 iff the edge does not contribute to the potential difference @xmath96 .
in particular , this implies that an edge is free if @xmath97 to construct the @xmath81-sets we use the probabilistic method .
more specifically , for every token @xmath98 and all nodes @xmath68 , we independently put @xmath99 with probability @xmath100 .
the following lemma shows that then only a small number of non - free edges are required in every graph @xmath60 .
[ lemma : freeedges ] if each set @xmath101 contains each token @xmath98 independently with probability @xmath100 , for every round @xmath57 and every token assignment @xmath102 , the graph @xmath103 induced by all free edges in round @xmath57 has at most @xmath21 components with probability at least @xmath104 .
assume that the graph @xmath103 has at least @xmath105 components for some @xmath106 .
@xmath103 then needs to have an independent set of size @xmath105 , i.e. , there needs to be a set @xmath107 of size @xmath108 such that for all @xmath109 , the edge @xmath89 is not free in round @xmath57 . using and the fact that @xmath110 for all @xmath68 and @xmath70 , an edge @xmath89 is free in round @xmath57 if @xmath111 and @xmath112 or if @xmath113 . to argue that @xmath105 is always small we use a union bound over all @xmath114 ways to choose a set of @xmath105 nodes and all at most @xmath115 ways to choose the tokens to be sent out by these nodes .
note that since two nodes sending out the same token induce a free edge , all tokens sent out by nodes in @xmath116 have to be distinct .
furthermore , for any pair of nodes @xmath117 there is a probability of exactly @xmath118 for the edge @xmath89 to be free and this probability is independent for any pair @xmath119 with @xmath120 because nodes in @xmath116 send distinct tokens .
the probability that all @xmath121 node pairs of @xmath116 are non - free is thus exactly @xmath122 . if @xmath123 ( assuming @xmath124 ) , the union bound @xmath125 is less than @xmath126 as desired .
this shows that there is a way to choose the sets @xmath101 such that the greedy adversary always chooses a topology in which the graph @xmath103 induced by all free edges has at most @xmath127 components .
based on lemma [ lemma : freeedges ] , the lower bound from @xcite now follows almost immediately .
[ thm : original ] in an always connected dynamic network with @xmath0 tokens in which nodes initially know at most @xmath79 tokens on average , any centralized token - forwarding algorithm takes at least @xmath128 rounds to disseminate all tokens to all nodes . by independently including each token with probability @xmath126 in each of the sets @xmath129
, we have that @xmath130 with probability at least @xmath104 ( for sufficiently large @xmath44 ) .
further , by lemma [ lemma : freeedges ] , with probability at least @xmath104 , we obtain sets @xmath129 such that the potential can only grow by @xmath21 in every round . hence , there exists set @xmath101 such that the initial potential is at most @xmath131 and in each round , the potential function does not grow by more than @xmath21 . as in the end
the potential function has to reach @xmath44 , the claim then follows .
we conclude our discussion of generalizations of the basic lower bound proof of section [ sec : original ] by showing two relatively simple results concerning partial token dissemination and a related problem .
[ thm : deltalb ] for any @xmath132 , suppose an always connected dynamic network with @xmath0 tokens in which nodes initially know at most @xmath133 tokens on average .
then , any centralized token - forwarding algorithm requires at least @xmath134 rounds to solve @xmath13-partial token dissemination .
the proof is analogous to the proof in section [ sec : original ] .
again , we construct the @xmath81-sets using the probabilistic method . here
, we include every token in every set @xmath101 with probability @xmath135 .
for sufficiently large @xmath1 , we then get that @xmath136 with probability at least @xmath104 .
a potential of at least @xmath137 is needed to terminate at time @xmath10 .
following the same proof as for , there exists @xmath81-sets such that in each round the potential increases by at most @xmath138 which implies a @xmath139 lower bound .
let us now consider an interesting special case where initially one node knows all the tokens .
in this situation , a simple way of applying coding for token dissemination is to use forward error correcting codes ( fec ) . from the @xmath0 tokens ,
such a code generates a large number of code words ( of essentially the same length as one of the message ) , so that getting any @xmath0 code words allows to reconstruct all the @xmath0 messages . any token dissemination algorithm
as described above takes at least @xmath140 rounds to disseminate @xmath0 tokens .
let @xmath10 be the time in which the fec - based token dissemination algorithm terminates .
the total number of different fec messages sent is at most @xmath10 and every node needs to receiver at least a @xmath141 fraction of these messages . from
we thus get that @xmath142 which leads to @xmath143 . as the network
can be a static network of diameter linear in @xmath1 , @xmath144 is clearly also a lower bound on the time needed .
together , the two bounds imply the claim of the theorem . in this section
we show that it is possible to extend the lower bound to the case where nodes can send out @xmath145 tokens in each round .
note that it is a priori not clear that this can be done as for instance the related @xmath39 lower bound of @xcite breaks down completely if nodes are allowed to send two instead of one tokens in each round . in order to prove a lower bound for @xmath145
, we generalize the notion of free edges .
let us first consider a token assignment for the case @xmath145 .
instead of sending a single token @xmath85 , each node @xmath68 now broadcasts a set @xmath146 of at most @xmath24 tokens in every round @xmath57 .
analogously to before , we call the collection of pairs @xmath147 for @xmath87 , the token assignment of round @xmath57 .
we define the weight of an edge in round @xmath57 as the amount the edge contributes to the potential function growth in round @xmath57 .
hence , the weight @xmath148 of an edge @xmath95 is defined as @xmath149 as before , we call an edge @xmath150 with weight @xmath151 free . given the edge weights and the potential function as in section [ sec : original ] , a simple possible strategy of the adversary works as follows .
in each round , the adversary connects the nodes using an mst w.r.t .
the weights @xmath148 for all @xmath152 .
the total increase of the potential function is then upper bounded by the weight of the mst . for the mst to contain @xmath153 or more edges of weight at least @xmath154
, there needs to be set @xmath116 of @xmath155 nodes such that the weight of every edge @xmath89 for @xmath109 is at least @xmath154 .
the following lemma bounds the probability for this to happen , assuming that the @xmath81-sets are chosen randomly such that every token @xmath98 is contained in every set @xmath101 with probability @xmath156 for some constant @xmath157 .
[ lemma : cliques ] assume that each set @xmath129 contains each token independently with probability @xmath158 .
then , for every token assignment @xmath159 , there exists a set @xmath116 of size @xmath155 such that all edges connecting nodes in @xmath116 have weight at least @xmath154 with probability at most @xmath160 consider an arbitrary ( but fixed ) set of nodes @xmath161 and a set of token sets @xmath162 ( we assume that the token assignment contains the @xmath155 pairs @xmath163 ) .
we define @xmath164 to be the event that @xmath165 .
note that whenever @xmath166 grows by more than @xmath167 , the event @xmath164 definitely happens . in order to have @xmath168 for each @xmath169 ,
at least @xmath170 of the events @xmath164 need to occur .
hence , for all edges @xmath171 , @xmath172 , to have weight at least @xmath154 , at least @xmath170 of the events @xmath164 have to happen . as the event @xmath164 only depends on the randomness used to determine @xmath173 ,
events @xmath164 for different @xmath98 are independent .
the number of events @xmath164 that occur is thus dominated by a binomial random variable @xmath174\big)$ ] variable with parameters @xmath155 and @xmath175 $ ] .
the probability @xmath176 $ ] for each @xmath98 can be bounded as follows : @xmath177 \leq { \ell b\choose \ell w/4 } \cdot \left(\frac{{\varepsilon}}{4e b}\right)^{\ell w/4 } \leq \left(\frac{4e\ell b}{\ell w}\right)^{\ell w/4}\cdot\left(\frac{{\varepsilon}}{4e b}\right)^{\ell w /4 } = \left(\frac{{\varepsilon}}{w}\right)^{\ell w /4}.\ ] ] let @xmath178 be the number of events @xmath164 that occur .
we have @xmath179 \leq { \ell+1\choose ( \ell+1)/3}\cdot \left(\frac{{\varepsilon}}{w}\right)^{\frac{\ell w}{4}\cdot \frac{\ell+1}{3 } } \leq 2^{\ell+1}\cdot \left(\frac{{\varepsilon}}{w}\right)^{\frac{\ell w}{4}\cdot \frac{\ell+1}{3}}.\ ] ] the number of possible ways to choose @xmath155 nodes and assign a set of @xmath24 tokens to each node is @xmath180 the claim of the lemma now follows by applying a union bound over all possible choices @xmath161 and @xmath162 . based on lemma [ lemma : cliques ]
, we obtain the following theorem .
[ thm : multipletoken ] on always connected dynamic networks with @xmath0 tokens in which nodes initially know at most @xmath79 tokens on average , every centralized randomized token - forwarding algorithm requires at least @xmath181 rounds to disseminate all tokens to all nodes . for @xmath182 ,
let @xmath183 be the size of the largest set @xmath184 , such that that edge between any two nodes @xmath185 has weight at least @xmath186 .
hence , in the mst , there are at most @xmath187 edges with weight between @xmath186 and @xmath188 .
the amount by which the potential function @xmath92 increases in round @xmath57 can then be upper bounded by @xmath189 by lemma [ lemma : cliques ] ( and a union bound over the @xmath190 different @xmath186 ) , for a sufficiently small constant @xmath157 , @xmath191 with high probability .
the number of tokens learned in each round can thus be bounded by @xmath192 by a standard chernoff bound , with high probability , the initial potential is of the order @xmath193 .
therefore to disseminate all tokens to all nodes , the potential has to increase by @xmath194 and the claim follows .
while allowing that the network can change arbitrarily from round to round is a clean and useful theoretical model , from a practical point of view it might make sense to look at dynamic graphs that are a bit more stable . in particular
, some connections and paths might remain reliable over some period of time . in @xcite , token dissemination and the other problems considered
are studied in the context of @xmath10-interval connected graphs . for @xmath10 large enough ,
sufficiently many paths remain stable for @xmath10 rounds so that it is possible to use pipelining along the stable paths to disseminate tokens significantly faster ( note that this is possible even though the nodes do not know which edges are stable ) . in the following , we show that the lower bound described in section [ sec : original ] can also be extended to @xmath10-interval connected networks . on @xmath10-interval connected dynamic networks in which nodes initially know at most @xmath79 of @xmath0 tokens on average , every randomized token - forwarding algorithm requires at least @xmath195 rounds to disseminate all tokens to all nodes .
we assume that each of the sets @xmath101 independently contains each of the @xmath0 tokens with probability @xmath196 for a sufficiently small constant @xmath157 . as before , we let @xmath85 be the token broadcast by node @xmath68 in round @xmath57 and call the set of pairs @xmath86 the token assignment of round @xmath57 . in the analysis , we will also make use of token assignments of the form @xmath197 , where @xmath198 is a set of tokens sent by some node @xmath68 .
given a token assignment @xmath199 , as in the previous subsection , an edge @xmath89 is free in particular if @xmath200 .
let @xmath201 be the free edges w.r.t . a given token assignment @xmath202 .
further , we define @xmath203 to be the partition of @xmath16 induced by the components of the graph @xmath204 .
consider a sequence of @xmath205 consecutive rounds @xmath206 .
for a node @xmath207 and round @xmath208 , @xmath209 $ ] , let @xmath210 be the set of tokens transmitted by node @xmath207 in rounds @xmath211 and let @xmath212 .
as above , let @xmath213 be the free edges for the token assignment @xmath214 and let @xmath215 be the partition of @xmath16 induced by the components of the graph @xmath216
. note that for @xmath217 , @xmath218 and @xmath219 is a sub - division of @xmath215 .
we construct edge sets @xmath220 as follows . the set @xmath221 contains @xmath222 edges to connect the components of the graph @xmath223 . for @xmath224 , the edge set @xmath225
is chosen such that @xmath226 , @xmath227 , and the graph @xmath228 is connected .
note that such a set @xmath225 exists by induction on @xmath98 and because @xmath215 is a sub - division of @xmath229 . for convenience ,
we define @xmath230 . by the above construction
, the number of edges in @xmath231 is @xmath232 , where @xmath233 is the number of components of the graph @xmath216 . because in each round , every node transmits only one token , the number of tokens in each @xmath234 is at most @xmath235 . by lemma [ lemma : cliques ] , if the constant @xmath236 is chosen small enough , the number of components of @xmath204 and therefore the size of @xmath231 is upper bounded by @xmath237 , w.h.p .
we construct the dynamic graph as follows . for simplicity , assume that the first round of the execution is round @xmath238 . consider some round @xmath57 and let @xmath239 be the largest round number such that @xmath240 and @xmath241 . the edge set in round @xmath98 consists of the the free edges in round @xmath98 , as well as of the sets @xmath242 and @xmath243 .
the resulting dynamic graph is @xmath10-interval - connected .
furthermore , the number of non - free edges in each round is @xmath244 . because in each round
, at most @xmath245 tokens are learned over each non - free edge , the theorem follows .
[ lem : randomgraph ] for any @xmath1 and constant @xmath132 an erdoes - renyi random graph @xmath246 is @xmath144-vertex - connected with probability at least @xmath247 .
do a union bound over all sizes @xmath105 and all cuts of size @xmath105 to have at least @xmath0 neighbors . rather than requiring more connectivity over time
, we now consider the case when the network is better connected in every round . if the network is @xmath7-vertex connected for some @xmath30 , in every round , each set of nodes can potentially reach @xmath7 other nodes ( rather than just @xmath64 ) . in @xcite
, it is shown that for the basic greedy token forwarding algorithm , one indeed gains a factor of @xmath38 if the network is @xmath7-vertex connected in every round .
we first need to state two general facts about vertex connected graphs .
[ lem : cconnect ] if in a graph @xmath63 there exists a vertex @xmath61 with degree at least @xmath7 such that @xmath248 is @xmath7-vertex connected then @xmath63 is also @xmath7-vertex connected .
[ lem : kconnectmindegree ] for @xmath7 , any @xmath1-node graph @xmath62 with minimum degree at least @xmath249 can be augmented by @xmath1 edges to be @xmath7-vertex connected .
we specialize the much more powerful results of @xcite which characterize the minimum number of augmentation edges needed to our setting : according to ( * ? ? ?
* p41 , criterion 4 ) any graph with minimum degree at least @xmath249 is @xmath7-independent and for such a graph @xmath63 it holds that the minimum number of edges needed to make it @xmath7-vertex connected is exactly @xmath250 ( * ? ? ?
* theorem 3.12 ) . here ,
@xmath251 is the maximum number of connected components @xmath63 can be dissected by removing @xmath252 nodes ( which is at most @xmath253 ) and @xmath254 is at most the maximum value for @xmath255 that can be obtained for a disjoint node partitioning @xmath256 ( * ? ? ?
* p33 ) . here
@xmath257 is the set of nodes neighboring @xmath258 , i.e. , @xmath259 .
because every node has degree at least @xmath249 , @xmath260 and thus @xmath261 we will also need the following basic result about weighted sums of bernoulli random variables .
[ lem : increase ] for some @xmath7 let @xmath262 be positive integers with @xmath263 .
furthermore , let @xmath264 be i.i.d . bernoulli variables with @xmath265 = { \mathbb{p}}[x_i = 0 ] = 1/2 $ ] for all @xmath98 . for any integer
@xmath266 it holds that : @xmath267\land \sum_{i \in \{j | x_j = 1\ } \cup l } \ell_i \geq c\right\ } } > x \right ] < 2^{-\theta(\frac{x\ell}{c})}.\ ] ] that is , the probability that @xmath268 of the random variables need to be switched to one after a random assignment in order get @xmath269 is at most @xmath270 .
fix a positive integer @xmath268 .
suppose without loss of generality that @xmath271 .
clearly @xmath272 always holds if @xmath273 .
thus , there is nothing to show unless @xmath274 for all @xmath275 and @xmath276 .
for this case , consider a scaling by a factor of @xmath277 of all the values .
the scaled values @xmath278 for @xmath279 are at most one and the scaled expectation of the sum is @xmath280 > \frac{2}{3}\ell\cdot\frac{x}{c}\cdot\frac{1}{2}=\frac{\ell x}{3c}$ ] .
a standard chernoff bound then shows that @xmath281 < 2^{-\theta(\frac{x\ell}{c})}$ ] . to prove our lower bound for always @xmath7-connected graph , we initialize the @xmath81-sets as for always connected graphs , i.e. , each token @xmath98 is contained in every set @xmath101 with constant probability @xmath282 ( we assume @xmath283 in the following ) . in each round
, the adversary picks a @xmath7-connected graph with as few free edges as possible . using lemmas [ lem : cconnect ] and [ lem : kconnectmindegree ]
, we will show that a graph with a small number of non - free edges can be constructed as follows .
first , as long as we can , we pick vertices with at least @xmath7 neighbors among the remaining nodes .
we then show how to extend the resulting graph to a @xmath7-connected graph .
[ lem : cconlb1 ] with high probability ( over the choices of the sets @xmath101 ) , for every token assignment @xmath159 , the largest set @xmath116 for which no node @xmath284 has at least @xmath7 neighbors in @xmath116 is of size @xmath285 .
consider some round @xmath57 with token assignment @xmath286 . we need to show that for any set @xmath116 of size @xmath287 for a sufficiently large constant @xmath288 , at least one node in @xmath116 has at least @xmath7 free neighbors in @xmath116 ( i.e. , the largest degree of the graph induced by the free edges between nodes in @xmath116 is at least @xmath7 ) .
we will use a union bound over all @xmath289 sets @xmath116 and all @xmath115 possibilities for selecting the tokens sent by these nodes .
we want to show that if the constant @xmath288 is chosen sufficiently large , for each of these @xmath290 possibilities we have a success probability of at least @xmath291 .
we first partition the nodes in @xmath116 according to the token sent out , i.e. , @xmath184 is the subset of nodes sending out token @xmath98 .
note that if for some @xmath292 we have @xmath293 we are done since all edges between nodes sending the same token are free . with this , let @xmath294 be such that @xmath295 and @xmath296 .
we now claim that for every @xmath297 , with probability at most @xmath298 , there does not exist a node in @xmath299 that has at least @xmath7 free edges to nodes in @xmath300 .
note that the events that a node from @xmath299 has at least @xmath7 free edges to nodes in @xmath301 are independent for different @xmath292 as it only depends on which nodes @xmath68 in @xmath301 have @xmath292 in @xmath101 and on the @xmath302-sets of the nodes in @xmath299 .
the claim that we have a node with degree @xmath7 in @xmath116 with probability at least @xmath303 then follows from the definition of @xmath294 .
let us therefore consider a fixed value @xmath292 .
we first note that for a fixed @xmath292 by standard chernoff bounds with probability at least @xmath304 , there at least @xmath305 nodes in @xmath301 that have token @xmath292 in their initial @xmath81-set . for @xmath288
sufficiently large , this probability is at least @xmath306 . in the following ,
we assume that there are at least @xmath307 nodes @xmath68 in @xmath301 for which @xmath308 .
let @xmath309 for any @xmath310 denote the number of nodes in @xmath184 that have token @xmath292 in the initial @xmath81-set .
the number of free edges to a node @xmath68 in @xmath299 is at least @xmath311 , where the random variable @xmath312 is @xmath64 if node @xmath68 initially has token @xmath98 in @xmath129 and @xmath238 otherwise ( i.e. , @xmath312 is a bernoulli variable with parameter @xmath20 ) .
note that since @xmath313 , the expected value of the number of free edges to a node @xmath68 in @xmath299 is at least @xmath314 . by a chernoff bound , the probability that the number of free edges from a node @xmath68 in @xmath299 does not deviate by more than a constant factor with probability @xmath315 .
note that @xmath316 since @xmath317 .
for @xmath288 large enough this probability is at least @xmath318 .
because the probability bound only depends on the choice of @xmath101 , we have independence for different @xmath319 .
therefore , given that at least @xmath307 nodes in @xmath301 have token @xmath292 , the probability that no node in @xmath299 has at least @xmath7 neighbors in @xmath301 can be upper bounded as @xmath320 .
together with the bound on the probability that at least @xmath307 nodes in @xmath301 have token @xmath292 in their @xmath302 set , the claim of the lemma follows .
lemma [ lem : cconlb1 ] by itself directly leads to a lower bound for token forwarding algorithms in always @xmath7-vertex connected graphs .
[ cor : simplecconn ] suppose an always @xmath7-vertex connected dynamic network with @xmath0 tokens in which nodes initially know at most a constant fraction of the tokens on average
. then , any centralized token - forwarding algorithm takes at least @xmath321 rounds to disseminate all tokens to all nodes . by lemma [ lem : cconlb1 ]
, we know that there exists @xmath302-sets such that for every token assignment after adding all free edges , the size of the largest induced subgraph with maximum degree less than @xmath7 is @xmath285 . by lemma [ lem : cconnect ] , it suffices to make the graph induced by these @xmath285 nodes @xmath7-vertex connected to have a @xmath7-vertex connected graph on all @xmath1 nodes . to achieve this , by lemma [ lem : kconnectmindegree ]
, it suffices to increase all degrees to @xmath249 and add another @xmath285 edges .
overall , the number of non - free edges we have to add for this is therefore upper bounded by @xmath322 .
hence , the potential function increases by at most @xmath322 per round and since we can choose the @xmath302-sets so that initially the potential is at most @xmath323 for a constant @xmath324 , the bound follows .
as shown in the following , by using a more careful analysis , we can significantly improve this lower bound for @xmath325 .
note that the bound given by the following theorem is at most an @xmath326 factor away from the simple `` greedy '' upper bound .
[ lem : cconlb2 ] suppose an always @xmath7-vertex connected dynamic network with @xmath0 tokens in which nodes initially know at most a constant fraction of the tokens on average .
then , any centralized token - forwarding algorithm takes at least @xmath327 rounds to disseminate all tokens to all nodes .
we use the same construction as in to obtain a set @xmath116 of size @xmath328 for a sufficiently large constant @xmath329 such that @xmath116 needs to be augmented to a @xmath7-connected graph .
note that we want the set to be of size @xmath105 and therefore we do not assume that in the induced subgraph , every node has degree less than @xmath7 .
we improve upon by showing that it is possible to increase the potential function by adding a few more tokens to the @xmath81-sets , so that afterwards it is sufficient to add @xmath330 additional non - free edges to @xmath116 to make the induced subgraph @xmath7-vertex connected .
hence , an important difference is that are not counting the number of edges that we need to add but the number of tokens we need to give away ( i.e. , add to the existing @xmath81-sets ) .
we first argue that w.h.p . , it is possible to raise the minimum degree of vertices in the induced subgraph of @xmath116 to @xmath331 without increasing the potential function by too much
. then we invoke and get that at most @xmath330 more edges are then needed to make @xmath116 induce a @xmath7-connected graph as desired .
we partition the nodes in @xmath116 according to the token sent out in the same way as in the proof of , i.e. , @xmath184 is the subset of nodes sending out token @xmath98 .
let us first assume that no set @xmath184 contains more than @xmath332 nodes .
we can then divide the sets of the partition into two parts with at least @xmath332 nodes each . to argue about the sets
, we rename the tokens sent out by nodes in @xmath116 as @xmath333 so that we can find a token @xmath294 for which @xmath334 and @xmath335 .
we call the sets @xmath299 for @xmath336 the left side of @xmath116 and the sets @xmath299 for @xmath337 the right side of @xmath116 .
if there is a set @xmath184 with @xmath338 , we define @xmath184 to be the right side and all other sets @xmath299 to be the left side of @xmath116 .
we will show that we can increase the potential function by at most @xmath339 such that all the nodes on the left side have at least @xmath331 neighbors on the right side .
if all sets @xmath184 are of size at most @xmath332 , increasing the degrees of the nodes on the right side is then done symmetrically .
if the right side consists of a single set @xmath184 of size at least @xmath332 , for @xmath288 large enough we have @xmath340 and therefore nodes on the right side already have degree at least @xmath331 by just using free edges .
we start out by adding some tokens to the sets @xmath341 for nodes @xmath68 on the right side such that for every token @xmath342 on the left side , there are at least @xmath343 nodes @xmath68 on the right side for which @xmath344 .
let us consider some fixed token @xmath342 from the left side .
because every node @xmath68 on the right side has @xmath308 with probability @xmath20 , with probability at least @xmath345 , at least @xmath346 nodes u on the right side have @xmath308 .
for such a token @xmath292 , we do not need to do anything .
note that the events that @xmath308 are independent for different @xmath292 on the left side
. therefore , for a sufficiently large constant @xmath347 and a fixed collection of @xmath348 tokens @xmath292 sent by nodes on the left side , the probability that none of these tokens is in at least @xmath343 sets @xmath101 for @xmath68 on the right side is at most @xmath349 for a given constant @xmath350 .
as there are at most @xmath105 tokens sent by nodes on the left side , the number of collections of @xmath348 tokens is at most @xmath351 which is less than @xmath352 for sufficiently large @xmath288 .
hence , with probability at least @xmath353 , for at most @xmath348 tokens @xmath292 on the left side there are less than @xmath343 nodes @xmath68 on the right that have @xmath354 . for these @xmath21 tokens
@xmath292 , we add to @xmath292 to @xmath341 for at most @xmath343 nodes @xmath68 on the right side , such that afterwards , for every token @xmath292 sent by a node on the left side , there are at least @xmath343 nodes @xmath68 on the right for which @xmath344 .
note that this increases the potential function by at most @xmath355 .
we next show that by adding another @xmath356 tokens to the @xmath81-sets of the nodes on the left side , we manage to get that every node @xmath68 on the left side has at least @xmath331 free neighbors on the right side . for a token @xmath342 sent by some node on the left side and a token @xmath310 sent by some node on the right side , let @xmath357 be the number of nodes @xmath358 for which @xmath344 .
note that if token @xmath98 is in @xmath359 for some @xmath360 , @xmath61 has @xmath357 neighbors in @xmath184 . using the augmentation of the @xmath341-sets for nodes on the right
, we have that for every @xmath342 , @xmath361 . for every @xmath310 , with probability @xmath20 , we have @xmath362 .
in addition , we add tokens additional @xmath98 to @xmath359 for which @xmath363 such that in the end , @xmath364 . by lemma [ lem :
increase ] , the probability that we need to add @xmath365 tokens is upper bounded by @xmath366 .
as the number of tokens we need to add to @xmath367 is independent for different @xmath61 , in total we need to add at most @xmath368 tokens with probability at least @xmath353 .
note that this is still true after a union bound over all the possible ways to distributed the @xmath356 tokens among the @xmath369 nodes .
using lemma [ lem : kconnectmindegree ] , we then have to add at most @xmath370 additional non - free edges to make the graph induced by @xmath116 @xmath7-vertex connected .
there are at most @xmath371 ways to choose the set @xmath116 and @xmath372 ways to assign tokens to the nodes in @xmath116 .
hence , if we choose @xmath373 sufficiently large , the probability that we need to increase the potential by at most @xmath374 for all sets @xmath116 and all token assignments is positive .
the theorem now follows as in the previous lower bounds ( e.g. , as in the proof of theorem [ thm : original ] ) .
we would like to thank chinmoy dutta , gopal pandurangan , rajmohan rajaraman , and zhifeng sun for helpful discussions and for sharing their work at an early stage .
a. f. anta , a. milani , m. a. mosteiro , and s. zaks .
opportunistic information dissemination in mobile ad - hoc networks : the profit of global synchrony . in _ proc .
24th int .
symp . on distributed computing ( disc ) _ ,
pages 374388 , 2010 . c. avin , m. kouck , and z. lotker . how to explore a fast - changing world ( cover time of a simple random walk on evolving graphs ) .
in _ proc . of 35th coll .
on automata , languages and programming ( icalp ) _ , pages 121132 , 2008 .
a. clementi , c. macci , a. monti , f. pasquale , and r. silvestri . flooding time in edge - markovian dynamic graphs . in _ proc . of 27th acm symp . on principles of distributed computing ( podc ) _
, pages 213222 , 2008 .
b. haeupler and m. mdard .
one packet suffices - highly efficient packetized network coding with finite memory . in _ information theory proceedings ( isit ) , 2011 ieee international symposium on _ , pages 1151 1155 , 2011 .
f. kuhn , c. lenzen , t. locher , and r. oshman .
optimal gradient clock synchronization in dynamic networks . in _ proc . of 29th acm symp . on principles of distributed computing ( podc ) _ , pages 430439 , 2010 . | we study lower bounds on information dissemination in adversarial dynamic networks .
initially , @xmath0 pieces of information ( henceforth called tokens ) are distributed among @xmath1 nodes .
the tokens need to be broadcast to all nodes through a synchronous network in which the topology can change arbitrarily from round to round provided that some connectivity requirements are satisfied .
if the network is guaranteed to be connected in every round and each node can broadcast a single token per round to its neighbors , there is a simple token dissemination algorithm that manages to deliver all @xmath0 tokens to all the nodes in @xmath2 rounds .
interestingly , in a recent paper , dutta et al . proved an almost matching @xmath3 lower bound for deterministic token - forwarding algorithms that are not allowed to combine , split , or change tokens in any way . in the present paper , we extend this bound in different ways . if nodes are allowed to forward @xmath4 tokens instead of only one token in every round , a straight - forward extension of the @xmath2 algorithm disseminates all @xmath0 tokens in time @xmath5 .
we show that for any randomized token - forwarding algorithm , @xmath6 rounds are necessary . if nodes can only send a single token per round , but we are guaranteed that the network graph is @xmath7-vertex connected in every round , we show a lower bound of @xmath8 , which almost matches the currently best @xmath9 upper bound .
further , if the network is @xmath10-interval connected , a notion that captures connection stability over time , we prove that @xmath11 rounds are needed .
the best known upper bound in this case manages to solve the problem in @xmath12 rounds .
finally , we show that even if each node only needs to obtain a @xmath13-fraction of all the tokens for some @xmath14 $ ] , @xmath15 are still required . |
[ sec : intro ] @xcite first discussed multiple regimes of turbulent chemical combustion and gave scaling relations for each . in modern terms , the two regimes can be distinguished by their karlovitz number , ka , ( e.g. , * ? ? ?
* ) , @xmath3 or equivalently , @xmath4 here @xmath0 is the _ rms _ velocity of turbulent fluctuations on an integral scale , @xmath5 ; @xmath6 is the laminar conductive speed ; @xmath7 is the width of the laminar flame ; and @xmath8 is the gibson length . for isotropic kolmogorov turbulence ( assumed throughout this paper ) ,
the turbulent speed on the scale of the flame thickness is @xmath9 and the gibson scale , the size of the eddy that turns over in a ( laminar ) flame crossing time , is @xmath10 for ka @xmath11 , individual laminar flames are moved around by the largest turbulent eddies while smaller eddies have little effect .
the overall burning progresses at a speed determined by turbulence properties and is independent of the burning rate on small scales .
this regime has been extensively explored in the astrophysical context @xcite and its properties are reflected in the munich group s subgrid model for flame propagation @xcite .
the condition ka @xmath12 , on the other hand , implies that turbulence can penetrate into the flame and transport heat , and possibly fuel , faster than laminar burning crosses a flame width .
an equivalent condition is that the gibson scale is much less than the flame thickness .
this regime too has been discussed in the astrophysical literature @xcite .
it is generally agreed that if spontaneous detonation is to occur , it requires ka @xmath13 and probably ka @xmath14 so that the burning region itself is disrupted , not just the preheat zone .
it is also known in the chemical combustion community @xcite that the region ka @xmath12 can be further divided based on the value of the damkhler number , da = @xmath15 . here
@xmath16 is the characteristic burning time scale , appropriately modified by turbulence . for da @xmath17 1 , the eddy turnover time on the integral scale
is short compared with the nuclear time ; for da @xmath18 1 , it is longer . in the literature ,
the term `` distributed reaction zone(s ) '' has been used with reference to the da @xmath17 1 regime , the da @xmath18 1 regime , or both lumped together ( i.e. , all flames with ka @xmath19 1 ) .
therefore we avoid this terminology , choosing instead to follow @xcite in referring to da @xmath17 1 as the `` well - stirred reactor regime '' ( wsr regime ) , and to follow @xcite in referring to da @xmath18 1 as the `` stirred flame regime '' ( sf regime ) . in the wsr regime , there is only one flame .
it has a width broader than the integral scale and a speed slower than the turbulent speed on the integral scale .
this sort of flame is similar to the usual laminar flame , except that the turbulent diffusion coefficient ( @xmath20 ) now substitutes for conduction . within the sf regime ,
on the other hand , there can be multiple burning regions , but the idea of a flame brush composed of individual flamelets with well - defined local properties is no longer valid .
the overall burning continues with an average rate given by the turbulent speed on the integral scale , but the flamelets do not have a uniform width and their number and individual speeds are quite variable . in this paper , we explore these three regimes of turbulent nuclear combustion , ka @xmath21 , wsr , and sf , in the context of a type ia supernova using a numerical tool , the linear eddy model ( [ sec : lem ] ) .
if a transition to detonation is to occur , we conclude that it must happen in the sf regime , specifically where da @xmath22 and in the presence of a high degree of turbulence , @xmath23 .
an example of a successful spontaneous transition to detonation is given .
[ sec : special ] the conditions characterizing turbulent ( nuclear ) combustion in a type ia supernova are novel and have no direct analogue on earth .
this makes the supernova an interesting environment for testing new physics , but it also means that our terrestrial intuition regarding flames can be misleading . for example , it is thought that laboratory flames in the ka @xmath12 regime simply go out @xcite because they are unable to maintain their heat in the presence of so much turbulence .
but the flame in a supernovae can never `` go out '' until the star comes apart and , in terms , of local flame variables , that takes a very long time .
the relevant time scale is the hydrodynamic time scale for the whole star , @xmath241 s , not the shorter local turbulent time scale , @xmath25 . also the star is very large , @xmath26 laminar flame thicknesses and 100 integral scales .
rare events have many opportunities for realization .
there are also other conditions that are vastly different and , in some cases , make detonation more easily achievable . the reynolds number in a type ia supernova is orders of magnitude greater than achieved thus far in any terrestrial experiment or in any numerical simulation .
the source of viscosity is electron - ion interactions in a fully ionized plasma @xcite @xmath27 for z = 7 , @xmath28 = 2 , @xmath29 , and @xmath30 = 0.5 .
the reynolds number is then @xmath31 for @xmath32 10@xmath33 g @xmath2 , @xmath0 = 100 km s@xmath34 , and @xmath5 = 10 km .
this large value of re implies a tiny kolmogorov scale , much smaller than any laminar flame thickness under consideration here , and irresolvably small in most numerical simulations , @xmath35 note also that this implies ten orders of magnitude in length scale where the turbulence is ( assumed to be ) kolmogorov and isotropic with constant energy dissipation , @xmath36 . on earth ,
the kolmogorov scale is usually not so small compared with the flame thickness .
this makes achieving the sf and wsr regimes more difficult on earth , though certainly not impossible .
[ sec : cp ] the heat capacity in the supernova at a relevant density , @xmath37 g @xmath2 , is due to a combination of semi - degenerate electrons , ions and radiation .
radiation is an important component of the heat capacity at these low densities and hence the heat capacity is a rapidly increasing function of the temperature . at a density of @xmath38 g @xmath2 , for 50% c and 50% o ,
the heat capacity at constant pressure is 3.5 , 6.0 , 9.0 , 12.7 , 18.2 , , and 27.3 @xmath39 erg gm@xmath34 k@xmath34 for a temperature of 0.5 , 1.0 , 1.5 , 2.0 , 2.5 , and 3.0 @xmath40 k respectively .
the power of t upon which @xmath41 depends varies from 0.70 to 2.70 in the same temperature range .
this means that when the fuel is cold , a small amount of burning raises the temperature a lot .
given the high power of temperature upon which the burning depends , burning just a little fuel dramatically shortens the nuclear time scale .
[ sec : enuct ] the most important reaction rate in the regime where detonation might occur is @xmath42c + @xmath42c .
the rate for this reaction is proportional to @xmath43 with @xmath44 for temperatures in the range 1 - 3 @xmath40 k ( more sensitive at lower temperature ) .
this high temperature sensitivity coupled with the temperature - dependent heat capacity means that once about half of the carbon has burned , the remaining increase in the temperature happens very rapidly . as we shall see later this leads to small but very rapid increases in the local pressure that can help initiate a detonation . as burning plumes of ash float due to the rayleigh - taylor instability , they create shear and turbulence on their boundaries
. since the plumes are large , the speed at which buoyancy balances drag is high .
a speed of order 10 - 30% sonic is necessary to burn a large fraction of the star before it comes apart .
typical turbulent speeds on an integral scale of 10 km are about 150 km s@xmath34 @xcite , but speeds as great as 1000 km s@xmath34 may occasionally occur . the sound speed in the star at a density of 10@xmath33 g @xmath2 is 3500 km s@xmath34 , so the fastest turbulence is not terribly subsonic .
fluctuations in burning rate do not have to accelerate the burning by orders of magnitude in order to make a detonation happen .
[ sec : turb ] the lewis number , which is the ratio of thermal diffusivity to mass diffusivity , is very large in the supernova .
the ionic diffusion coefficient for a carbon - oxygen plasma is @xcite @xmath45 where @xmath46 is the plasma frequency , @xmath47 , with @xmath48 , the ion density , and @xmath49 .
the thermal diffusion coefficient is @xmath50 here representative conditions and opacities have been assumed : @xmath51 k , @xmath52 g @xmath2 , and @xmath53 @xmath54 gm@xmath34 @xcite . combining these two equations , @xmath55 .
terrestrial lewis numbers are close to unity .
the large lewis number in the star has an effect on the laminar speed ( [ sec : laminar ] ) , but this dependence is greatly mitigated in the turbulent regime ( [ sec : distrib ] ) where the turbulent diffusion coefficient exceeds @xmath56 . in that case , the effective lewis number approaches unity since both ions and heat are transported with equal efficiency by the turbulent eddies @xcite .
[ sec : lem ] the range of scales that must be resolved to address the flame propagation problem in a type ia supernova is very large , @xmath57 cm even if the kolmogorov scale is not resolved .
this range exceeds the current or anticipated capabilities of 3d simulations , or even 2d simulations .
this conundrum arises for many turbulent flow environments , and motivated the development of the linear eddy model ( lem ) , a 1d simulation tool @xcite .
lem simulates the evolution of scalar properties on a 1d spatial domain .
this can be interpreted as property profile evolution along a 1d line of sight through 3d turbulent flow .
the 1d domain is treated as a closed system with respect to enforcement of conservation laws . in this and other respects
, lem is not fully consistent with the evolution observed along a line of sight , yet it captures the salient features and provides useful results . the physical processes that are time advanced on the lem domain are diffusive ( e.g. , fickian ) transport ( in the present context representing species transport , radiation transport or subgrid turbulent transport ) , chemical ( or nuclear ) reactions , and turbulent eddy motions . for combustion applications , including the present application , lem includes an equation of state , an energy equation , and thermal expansion , using a zero - mach - number ( constant - pressure ) formulation @xcite .
the novelty of the lem approach is the representation of eddy motions . on a 1d domain ,
advection in the usual sense can not reorder the fluid elements along the domain , hence can not emulate the folding of material surfaces that is an essential feature of turbulent stirring .
a model of eddy motions is introduced that is formulated to capture this folding effect and other properties of turbulent stirring .
namely , a turbulent eddy is represented as an instantaneous map , called the triplet map , that is applied to a designated interval of the 1d domain .
the triplet map is applied by first compressing all property profiles in the interval by a factor of three .
the property profiles in the interval are then replaced by three side - by - side copies ( ` images ' ) of the compressed profiles .
the middle copy is then flipped .
this procedure preserves the continuity of spatial property profiles .
it also leaves the total linear measure ( model analog of volume ) of fluid corresponding to a given state or set of states ( based on the property values ) unchanged .
this is the 1d analog of the solenoidal condition .
flipping of the middle copy introduces fluid - element reordering analogous to the folding of material surfaces .
the three - fold compression emulates gradient amplification by compressive strain .
the dispersive effect of extensive strain is represented in lem by the mapping of neighboring fluid elements to different images .
computationally , the triplet map is implemented as a permutation of the cells of a uniform discretization of the 1d domain .
this and other details of lem are explained elsewhere ( kerstein 1991 ) . here , features relevant to what follows are described .
some symbols that are used have different meanings than in later sections .
an lem simulation time advances processes other than advection until it is time to implement a triplet map .
after the map , advancement of the other processes resumes until it is time for the next map .
model inputs are the initial and boundary conditions , the transport coefficients and rate constants governing diffusive and chemical advancement , and parameters controlling the time sequence of triplet maps .
map size @xmath58 is sampled from a probability density function ( pdf ) @xmath59 that is designed to reproduce relevant features of the inertial - range turbulent cascade .
one key feature is the turbulent diffusivity @xmath60 associated with maps of size @xmath61 .
the inertial - range scaling @xmath62 is enforced . in @xcite , it is shown that this implies @xmath63 , where @xmath64 is a normalization factor .
map size @xmath58 is restricted to the range @xmath65 $ ] , where @xmath66 and @xmath5 are the model analogs of the kolmogorov scale and the integral scale , respectively .
then as shown in the appendix , the total turbulent diffusivity is given by @xmath67 , where @xmath68 is the total frequency of maps of all sizes per unit domain length . here
, homogeneous turbulence is assumed , so none of the model parameters depend on location , and map location is sampled uniformly within the notionally infinite 1d domain .
the parameters @xmath66 , @xmath5 , and @xmath60 are model inputs and @xmath68 is determined from them .
@xmath60 is not typically known for turbulent flows , but rather is inferred from the relation @xmath69 , where @xmath70 is an empirical coefficient . here ,
lem results are compared to 3d simulations of turbulent premixed combustion for which @xmath0 and @xmath5 are known , but @xmath60 is not , so @xmath70 must be specified in order to evaluate the lem input parameter @xmath60 . @xcite
calibrated @xmath70 by comparing lem results for turbulent premixed methane - air combustion ( using a simplified chemical mechanism ) to turbulent burning velocity measurements .
they chose the value @xmath71 for some cases and @xmath72 for others , reflecting the wide variation of turbulent burning velocity results obtained in different experiments .
the experimental set - ups did not correspond to the idealized case of flames freely propagating through stationary , homogeneous turbulence . here
, lem results are compared to 3d simulation results more closely analogous to the lem flow configuration .
the @xmath70 value inferred from this comparison is close to the lower of the two @xmath70 values reported by smith and menon .
[ sec : modifications ] the standard ( i.e. , terrestrial chemical combustion ) version of lem was modified to use an equation of state , opacities , and a nuclear reaction network appropriate to the supernova .
the equation of state ( the helmholtz eos
* ) included pressure and energy contributions from radiation , ions ( treated as an ideal gas ) , and electrons and pairs of arbitrary speed and degeneracy .
the opacity routine @xcite included heat transport by radiative diffusion and conduction .
energy generation and composition changes were followed using a nuclear reaction network with 7 species : @xmath73he , @xmath42c , @xmath74o , @xmath75ne , @xmath76 mg , @xmath77si , and @xmath78ni ( though no appreciable @xmath78ni was ever produced in this study ) .
these species were connected together by a chain of @xmath79 reactions as well as the heavy ion reactions 3@xmath80 , @xmath42c + @xmath42c , @xmath74o + @xmath74o , and @xmath42c + @xmath74o . though it is an unpublished version written by the authors , the network was similar to that of @xcite .
the network did not include electron screening corrections to the reaction rates which were , in general , small at the low densities considered here .
testing the fast 7 isotope network against a larger 19 isotope network that included screening and about 85 reactions @xcite showed that the small network gave an energy generation that was typically 20 - 40% smaller . since this is equivalent to a small error in the temperature , the results of the 7 isotope , unscreened network were deemed sufficiently accurate .
the code was then checked and calibrated against several other studies of flame propagation . for simple laminar flames ( [ sec : laminar ] ) , the turbulence parameter , @xmath70 ( [ sec : lem ] ) , does nt enter .
for turbulent flames in the flamelet regime , the default setting @xmath81 was used ( [ sec : flamelet ] ) . however , for flames in the wsr regime ( [ sec : distrib ] ) , good agreement with prior 3d studies required a smaller value of @xmath82 .
this value is within the range anticipated from terrestrial experiments ( [ sec : lem ] ) and was used for all calculations in the wsr and sf regimes ( i.e. , all studies with ka @xmath83 ) unless otherwise noted .
future 3d studies , especially of the sf regime , are encouraged in order to gain better confidence and understanding of @xmath70 for flames of this sort .
[ sec : laminar ] in the absence of turbulence , a flame has a simple structure in which a self - similar profile of temperature and fuel concentration propagates into the fuel with a well - defined width and speed .
heat is transported by a combination of conduction and radiative diffusion , and this heat raises the temperature to the point where fuel burns on a diffusion time scale . such flames are well understood @xcite , even in the supernova context @xcite . in multiple dimensions , laminar flames may become deformed and take on a cusp - like appearance due to the landau - darrieus instability @xcite , but this instability does not lead to a major destabilization of the overall burning @xcite , nor does it greatly affect the propagation speed @xcite . as a test of the lem implementation and to provide a calibration point for more complex studies to follow ,
the one - dimensional laminar speed was calculated for carbon - burning flames at a variety of densities for two initial carbon mass fractions , @xmath84 = 0.50 and 0.75 .
the results are given in table 1 and fig .
[ fig : flame ] .
the speed of the flame with carbon mass fraction 0.50 and density 10@xmath33 g @xmath2 was @xmath85 cm s@xmath34 , in good agreement with the @xmath86 cm s@xmath34 found by @xcite , who used somewhat different nuclear physics .
the ash temperatures for the two compositions considered here were also in good agreement with the values calculated by @xcite .
the flame speeds for other densities were in agreement with previous studies by @xcite , and @xcite . in these calculations
, heat was transported by conduction and radiation and the laminar scale was well resolved .
zoning for the case of 10@xmath33 g @xmath2 , @xmath84 = 0.50 was 0.0244 cm and the flame width , several cm .
a recalculation of the flame speed with coarser grids gave essentially the same answer .
for resolutions of 0.196 cm and 0.0978 cm the calculated flame speed was 3.2 and 3.1 @xmath87 cm s@xmath34 , respectively , indicating that our calculation and that of @xcite were both converged and did not differ because of resolution .
some dependence on lewis number was noted .
two calculations of the speed for @xmath84 = 0.50 and le = 100 and le = 1000 gave speeds of @xmath88 cm s@xmath34 and @xmath89 cm s@xmath34 respectively .
[ sec : regime ] [ sec : flamelet ] the interior of a type ia supernova is turbulent , first because of the convection that precedes the explosion , which gives a lower bound for @xmath90 50 - 100 km s@xmath34 @xcite , and second because of the rayleigh - taylor and kelvin - helmholtz instabilities associated with the flame itself @xcite .
because these speeds are so much greater than the laminar speed ( table 1 ) , especially at low density , it is expected that turbulence has a major effect on flame propagation .
initially , however , the flame is very thin compared to the gibson length and is just carried around by the eddies @xcite .
overall the burning rate is independent of the speed of each little flamelet , and is governed instead by the speed of the largest turbulent eddies .
this is damkhler s `` large scale turbulence '' limit . to illustrate burning in this regime
, we considered a laminar flame similar to that in fig .
[ fig : flame ] with a carbon mass fraction of 0.75 and a fuel density @xmath91 g @xmath2 , but embedded in a turbulent background with characteristic speed 400 m s@xmath34 on an integral scale of 10 m. on this length scale in a supernova , one would actually expect much larger turbulent speeds , 10 to 100 km s@xmath34 , cascading down from still higher values on larger scales . for the time being such large
, highly turbulent regimes with narrow flames remain out of computational reach , even in 1d .
conditions chosen here were thus artificial , but illustrative .
the gibson scale here ( eq . [ eq : gibson ] ) is 21 cm , considerably larger than the laminar flame width , which is @xmath241 cm ( fig .
[ fig : flame ] ) , but still about still 50 times less than the integral scale .
the results are given in fig .
[ fig : flamelet ] . sometimes only a single flame was observed , but at the particular moment sampled here , there were eleven ( counting the small blip of entrained ash at 800 cm ) .
[ fig : flameletspd ] shows the overall propagation speed of the flame brush . here , as in the rest of the paper , the burning `` speed ''
is defined by an integral over the grid of the burning rate , @xmath92 where @xmath93 is the nuclear energy generation rate on the grid ( erg g@xmath34 s@xmath34 ) , and @xmath94 is the energy released by burning a gram of fuel . for fuel with 50% carbon ,
@xmath94 is @xmath95 erg g@xmath34 and for 75% carbon it is @xmath96 erg g@xmath34 . for the case shown in fig . [
fig : flameletspd ] , the average burning speed is 500 m s@xmath34 , close to the turbulent _ rms _ speed on the integral scale , 400 m s@xmath34 . in lem ,
one expects for the flamelet regime that @xmath97 = 18 @xmath98 ( see appendix ) . combining this with @xmath99 ( [ sec : lem ] ) and using the same value for @xmath70 as the simulation ( @xmath81 )
, one has @xmath100 , which is excellent agreement .
[ fig : flameletspd ] also shows that the burning rate is far from regular .
speeds over three times the average are sometimes seen . at other times
, the flame briefly almost goes out .
a similar behavior would be expected in the actual supernova early on when the density is higher and the turbulent speed is not that much greater than the laminar one
. then , as here , there would be just a few flames within the integral scale and the burning rate would be highly irregular .
for example , at a density of @xmath101 g @xmath2 with a carbon mass fraction of 0.50 , the laminar flame speed is 36 km s@xmath34 @xcite and the ratio @xmath102 is , for @xmath103 km s@xmath34 , also about 50 .
the difference at these high densities is that the flame would be irresolvably small , even in the present version of lem , for a calculation that carried the entire integral scale .
the flames at such high density are individually very thin and the laminar speed far below sonic , so detonation at early times is avoided . as the density declines , the gibson length shrinks and the flame brush contains an increasingly large number of flamelets .
just before entering the sf regime in the supernova , there are roughly a thousand flamelets within the flame brush ( @xmath104 km s@xmath34 ; @xmath105 m s@xmath34 ) .
the statistical variations in overall speed are therefore small . .
the speed , as measured by the overall energy generation on the grid , is highly variable ranging from a lower bound given by the laminar speed of a single flamelet ( @xmath106 cm s@xmath34 ) to several times the mean value .
the dashed line indicates
the average during the time interval studied , @xmath107 cm s@xmath34 , which is close to the _ rms _ turbulent speed on the integral scale , @xmath108 cm s@xmath34 .
the turnover time on the integral scale is 0.025 s and one sees the effect of the occasional large eddy in accelerating the burning .
[ fig : flameletspd],scaledwidth=47.5% ] as the density in the star declines below several times 10@xmath33 g @xmath2 , the gibson length becomes less than the laminar flame width and the transport of fuel and heat by turbulent advection starts to become important on the scale of the flame . by the time the density reaches @xmath91 g @xmath2 , for typical turbulent speeds , not only is the preheat region of the flame being mixed by turbulence , but the burning zone itself has been disrupted and combustion has a qualitatively different character @xcite . in the star , this transition is brought about by the slowing and thickening of the laminar flame with decreasing density in a background of turbulence with nearly constant energy dissipation rate and integral length scale .
[ sec : distrib ] before discussing the stirred flame regime further though ( [ sec : stirred ] ) , it is helpful to consider a limiting case where the turbulence thickens a single flame to a dimension greater than the integral scale , i.e. , @xmath109 , where @xmath58 is some integral scale to be varied subject to the condition @xmath110 = constant , where @xmath111 is the velocity on the integral scale , @xmath58 . while probably not realized in the supernova because of the large integral scale , this is the regime most easily accessible to multi - dimensional simulations and corresponds to what is commonly meant by the `` well - stirred reactor '' ( e.g. , fig . 7 of * ? ? ?
the solutions here also obey scaling relations predicted by @xcite .
g @xmath2 , integral scale , @xmath112 cm , and turbulent speed , @xmath113 cm s@xmath34 .
the purple and blue curves are the average temperature and carbon mass fraction from the 3d study of @xcite .
the black and red curves are from the corresponding study using lem and @xmath114 ( table 2 ) .
the speed of the flame calculated in 3d was @xmath115 cm s@xmath34 . with lem it was @xmath116 cm s@xmath34 , about five times the laminar flame speed for these conditions .
the flame widths calculated in the two studies are quite similar .
differences in the temperature are attributable , in part , to the different fuel temperature employed in the two studies - @xmath117 k for the 3d study and @xmath118 k for the lem study and the smaller network used for the 3d study ( see text ) .
[ fig : andywidth],scaledwidth=47.5% ] [ sec : aspden ] @xcite recently carried out 3d simulations of flame propagation in the wsr regime for a density and composition appropriate to type ia supernovae .
we focus here on their calculation at 10@xmath119 g @xmath2 , @xmath120 = 0.50 , and @xmath121 erg g s@xmath34
. figures [ fig : andywidth ] , [ fig : andy ] , [ fig : andyspd ] , and table 2 show the results for an lem calculation at the same values of integral scale , @xmath58 = 15 cm , and @xmath122 = 2.47 km s@xmath34 that they employed .
a single broad flame propagates at a steady speed about 5 times faster than the laminar value .
the width of the flame , as measured by the full width at half maximum of the energy generation curve is about 50 cm , or 20 times the laminar width ( fig .
[ fig : flame ] ) . the most noticeable differences with the 3d calculation ( fig .
[ fig : andywidth ] ) are a consequence of differing assumptions regarding the background ( fuel ) temperature in the supernova .
aspden et al .
used @xmath123 k ; here we use @xmath118 k. the lower temperature would characterize a region of low turbulence following some expansion prior to a detonation transition .
the latter is characteristic of the convection zone in the supernova at the time of runaway . for a non - turbulent medium ,
the former is more physical .
however , anticipating that we will be interested here in turbulent energies @xmath124 erg g@xmath34 s@xmath34 and that at least a few turnovers on the integral scale are expected prior to a detonation , the latter is also a reasonable assumption .
for @xmath125 cm s@xmath34 and @xmath126 km , 10@xmath74 erg g@xmath34 would be deposited in one turnover time . at a density of 10@xmath33 g @xmath2
, this corresponds to a temperature increase from @xmath127 k ( or essentially zero ) to @xmath118 k. because of the temperature sensitivity of the heat capacity ( [ sec : cp ] ) , we expect our results to be insensitive to the assumed fuel temperature .
however we do note that , if a fuel temperature of @xmath127 is assumed in lem , better agreement with the width is obtained than in fig .
[ fig : andywidth ] , but the best value of @xmath70 is lowered to @xmath128 . while lower values of @xmath70 are actually more favorable to detonation , we prefer to use the larger value because it is more consistent with the range seen in terrestrial experiments ( [ sec : lem ] ) and because a hotter fuel temperature is both physically justifiable and more stable numerically .
= 0.147 cm ) .
the maximum energy generation in both plots is normalized to @xmath129 erg g@xmath34 s@xmath34 .
the calculation in the first frame used a diffusive energy transport coefficient based on radiation and conduction .
the one in the second frame , virtually identical in result , used the subgrid model discussed in the text .
[ fig : andy],title="fig:",scaledwidth=47.5% ] = 0.147 cm ) .
the maximum energy generation in both plots is normalized to @xmath129 erg g@xmath34 s@xmath34 .
the calculation in the first frame used a diffusive energy transport coefficient based on radiation and conduction .
the one in the second frame , virtually identical in result , used the subgrid model discussed in the text .
[ fig : andy],title="fig:",scaledwidth=47.5% ] as a function of time .
the turnover time on the integral scale is @xmath130 s. it takes many turnover times to reach the terminal speed , but then that speed is maintained rather precisely .
the dashed line shows the average speed , @xmath116 cm s@xmath34 , or 5 times the laminar value .
[ fig : andyspd],scaledwidth=47.5% ] for the calculation shown in the first frame of fig .
[ fig : andy ] , energy transport by conduction and radiative diffusion was included , just as in the laminar flame cases , and zoning was sufficient to resolve a laminar flame had one been present . however , the transport here is really dominated by the turbulence .
the radiative diffusion coefficient ( eq . [ eq : drad ] ) varies from 730 @xmath54 s@xmath34 in the fuel to @xmath131 @xmath54 s@xmath34 in the ash while the turbulent diffusion coefficient ( [ sec : lem ] ; @xmath132 with @xmath133 km s@xmath34 , @xmath58 = 15 cm , and @xmath82 ) is @xmath134 @xmath54 s@xmath34 .
[ fig : c12oft ] shows the distribution of carbon with temperature in the case of a laminar flame ( fig .
[ fig : flame ] ) and the turbulent flame ( fig .
[ fig : andy ] ) .
both plots show a single - valued , monotonic relation . for the laminar flame
this is not surprising . at each temperature in the self - similar front
there is a unique carbon abundance .
when eddies , coupled with diffusion move heat and carbon around in a way that , in lem at least , is discontinuous , the result is perhaps more surprising .
it is also important that the curves are substantially different in the two cases . the curve for the turbulent case in fig .
[ fig : andy ] is what one would result were carbon to burn at constant pressure _ with no transport whatsoever _
this means the process that is transporting heat is transporting composition with equal efficiency , i.e. , the effective lewis number is one .
one expects this sort of behavior when turbulence is dominating in the transport of both .
the lewis number in the calculation was still 1000 , but small scale eddies kept the mixture so well stirred that even a small amount of ionic diffusion at the grid scale gave the same result as if le = 1 .
in fact , the actual turbulent transport would have been even more efficient had the resolution been higher .
recall that the kolmogorov scale is not resolved here .
if the turbulent diffusion coefficient , @xmath135 , is so much greater than conduction and radiation , then the results should be independent of @xmath56 .
the second frame of fig . [ fig : andy ] shows that this is indeed the case . here
the calculation sets conduction and radiative transport to zero , but instead uses a diffusion coefficient coupling individual zones of @xmath136 where @xmath137 is the grid resolution ( constant in this paper ) , and @xmath138 the subgrid diffusion coefficient @xmath139 represents the mixing effects of unresolved eddies smaller than the grid scale
. the results for @xmath82 are identical .
the exact value of @xmath139 is not very important so long as it is derived from a length scale that is very much less than the ( turbulently broadened ) flame thickness .
there is a characteristic eddy size that is chiefly responsible for the diffusion ( appendix a ) and so long as that scale is well resolved , the results are similar .
however there is a choice for @xmath139 that works best as the resolution becomes coarser .
[ fig : subgrid ] shows that a diffusion coefficient derived from the turbulent speed on the grid scale has much better scaling properties than one based on e.g. , @xmath140 , even though the smallest eddy in lem is @xmath140 .
the figure also shows that the diffusive properties of a temperature discontinuity remain unaltered as the resolution is decreased by a factor of 50 . , but still in the wsr regime .
the integral scales are 120 cm ( first frame ) and 960 cm ( second frame ) , and the corresponding _
rms _ turbulent speeds on those scales , 4.93 km s@xmath34 and 9.86 km s@xmath34 .
the average propagation speeds are 0.61 km s@xmath34 and and 2.3 km s@xmath34 .
note that both the flame speeds and widths scale as @xmath141 in this regime , both in this figure and as compared with fig .
[ fig : andy ] . in the flame with the larger integral scale ( right frame ) ,
one begins to see some structure in the flame and small `` ledges '' of mixture with nearly constant fuel abundance and temperature ( [ sec : ledges ] ) .
[ fig : andy2],title="fig:",scaledwidth=47.5% ] , but still in the wsr regime .
the integral scales are 120 cm ( first frame ) and 960 cm ( second frame ) , and the corresponding _
rms _ turbulent speeds on those scales , 4.93 km s@xmath34 and 9.86 km s@xmath34 .
the average propagation speeds are 0.61 km s@xmath34 and and 2.3 km s@xmath34 .
note that both the flame speeds and widths scale as @xmath141 in this regime , both in this figure and as compared with fig .
[ fig : andy ] . in the flame with the larger integral scale ( right frame ) ,
one begins to see some structure in the flame and small `` ledges '' of mixture with nearly constant fuel abundance and temperature ( [ sec : ledges ] ) .
[ fig : andy2],title="fig:",scaledwidth=47.5% ] [ sec : scaling ] encouraged by these results , which imply that one need not resolve the laminar scale in this regime in order to obtain accurate results , we proceeded to explore flame properties for a large range of @xmath58 consistent with the condition @xmath142 erg g@xmath34 s@xmath34 ( table 2 ) . in general , the speed of a flame propagated by diffusion should obey the scaling relation @xmath143 where @xmath144 .
[ fig : andy2 ] shows the effect on the flame of increasing @xmath58 at constant turbulent energy dissipation @xmath145 , mass density , and carbon mass fraction .
for awhile , the profiles follow the scaling implied by eq .
[ eq : scale ] .
if the burning time is independent of @xmath58 , @xmath146 for given , @xmath58 , the flame speed also remains roughly steady .
however , this scaling also predicts its own breakdown . since @xmath147 increases as @xmath141 while @xmath111 only increases as @xmath148 there must be a critical integral scale @xmath149 for which the two become equal .
@xmath149 is determined by the relation @xmath150 , where @xmath111 is the velocity fluctuation at scale @xmath149 . rewriting this as @xmath151 gives @xmath152 at @xmath153 , the flame width and integral scale are the same and the burning speed is approximately equal to the turbulent speed on that scale @xcite .
this is also the size of an eddy that burns in one turnover time .
the quantity , @xmath154 , is a constant by assumption .
approximate values of @xmath149 have been tabulated for various densities , turbulent energies and carbon mass fractions by @xcite .
better values for the cases considered here can be inferred from the data in table 2 which uses our reaction network .
once @xmath155 , the flamelets no longer has the appearance of a diffusively broadened structure , though the individual flamelets can occasionally coalesce into larger structures .
this is the sf regime of the next section . as fig .
[ fig : andy3 ] and fig .
[ fig : andyvofl ] show , the flame structure and speed both become highly variable as the integral scale approaches this value .
burning in the wsr , @xmath156 , produces flames that , for a given constant density and fuel composition , have a well - defined nuclear time scale , @xmath157 .
[ fig : taunuc ] shows the temperature , and nuclear time scale as function of carbon mass fraction in the fuel for two initial carbon concentrations ( 50% and 75% by mass ) and fuel density 10@xmath33 g @xmath2 .
the burning is assumed to be isobaric and two time scales are computed .
one time scale reflects the instantaneous rate of carbon consumption , @xmath158 ( the dash - dotted line ) .
more relevant to the flame speed , however , is the _ induction _ time scale , which is the time required to consume most of the remaining fuel . in a situation where burning increases the temperature , and therefore the reaction rate ,
the induction time can be much shorter than the instantaneous burning time scale ( defined by the current abundance divided by the current burning rate ) .
dividing the turbulent flame width by the turbulent flame speed for those cases in table 2 where both quantities are well defined gives an approximately constant value @xmath159 s for @xmath160 erg g@xmath34 s@xmath34 and @xmath120 = 0.50 .
this corresponds to a carbon mass fraction in fig .
[ fig : taunuc ] of @xmath161 where the energy generation is about 1/@xmath162 of its maximum . defining @xmath157 in this fashion as the induction time scale from the point where the energy generation reaches @xmath163 of its eventual maximum allows the computation of @xmath164 . for the case of carbon mass fraction in the fuel
= 75% , the induction time scale was calculated based on a starting carbon mass fraction of 37% .
the values of @xmath149 so determined are given in table 3 .
these values are about a factor of two less than given in @xcite because they refer to a time scale derived from the width of the energy generating region .
if one instead uses the larger width based upon temperature or carbon mass fraction , the derived values for @xmath149 are about twice as large , consistent with @xcite .
[ sec : stirred ] the stirred flame ( sf ) regime , which is characterized by ka @xmath12 and da = @xmath165 , is the most complex of the three regimes of turbulent burning . in this case
, there is no well - determined scale for the flame width .
structures of size @xmath166 persist to some extent , but since the flame experiences a new eddy of length @xmath149 , in approximately the same time it takes to burn a distance @xmath149 , it is subject to continual disruption . at times
there may be almost no burning ; at others , nearly simultaneous burning happens on scales much larger than @xmath149 .
the fact that the flame has no persistent steady state also reflects a poorly defined nuclear time scale .
each temperature has a different burning time scale and thus the characteristic widths of mixtures prepared by the turbulence is very temperature sensitive .
cooler mixtures have larger length scales . in general
though , the average time scales are longer and the flame structures larger than expected from our studies of the wsr . because of the large integral scale in the supernova
, the sf regime is encountered as soon as the karlovitz number exceeds about 10 .
the damkhler number at that point is already much greater than unity and the wsr regime ( da @xmath21 ) is encountered much later , if ever . [ sec : transition ] just before the laminar flame is disrupted at ka @xmath24 10 , there are @xmath167 flame surfaces folded in the flame brush .
the average spacing between these flamelets is @xmath168 ( table 3 ) and the thickness of each is @xmath7 . a short time later , as ka continues to increase , these flamelets are smeared out to make a smaller number of broader , faster structures with characteristic average thickness @xmath149 .
their number then is @xmath169 , and their spacing is @xmath170 .
so long as da @xmath12 , as it is at the transition , the new broadened structures will , on the average , still be separated by distances much greater than their size and will not coalesce into one large mixture .
the quantity @xmath149 ( eq . [ eq : lambda ] ) that plays a such a critical role in this discussion can be thought of as a generalization of the gibson length ( eq .
[ eq : gibson ] ) . for the flamelet regime ,
the gibson scale is the size of the eddy that would be crossed by a laminar flame with speed @xmath6 on an eddy turnover time .
it is smaller for larger turbulent energy .
the quantity @xmath149 , on the other hand , is also the size of an eddy that turns over on a nuclear time scale , but its size depends only on turbulence properties , not the radiative diffusion that sets @xmath6 .
it is larger for larger turbulent energy .
as the supernova expands , the gibson scale shrinks as @xmath171 declines . eventually when ka @xmath172 , @xmath173 , but except for differences due to a changing nuclear time scale ( [ sec : scaling ] ) , at ka = 1 , @xmath149 is also approximately equal to @xmath8 and @xmath7 .
thereafter , as the density decreases more , @xmath8 ceases to have much meaning since laminar flame speeds are no longer relevant .
the new relevant scale , @xmath174 , on the other hand , grows rapidly at lower density due to the increasing nuclear time scale .
what this means is that the transition from laminar burning to the sf regime is probably smooth and uneventful .
flamelets will grow gradually in speed and width as the density declines , not discontinuously .
because the nuclear time , and thus @xmath149 , lack precise definitions in the sf regime , and because of intermittency , one can not completely rule out large scale transient mixing at the laminar - sf transition , especially if the turbulent energy is very large , but a detonation here seems unlikely . if it did happen at such high density , a very bright supernova would result due to the near complete incineration of the star . [ sec : complex ] as the density continues to decrease , @xmath149 increases and there are fewer , thicker , faster turbulent flamelets in the flame brush ( fig . [
fig : andy3 ] and fig .
[ fig : andyvofl ] ) .
the limit of one flame of size @xmath175 is reached when da = 1 .
this is the condition suggested by @xcite as likely for detonation .
the present study shows , however , that mixed regions larger than @xmath149 can exist transiently even for da @xmath12 .
since the critical mass for detonation decreases rapidly with increasing density , this makes detonation easier .
consider the case of 50% carbon , a turbulent dissipation rate of 10@xmath176 erg g@xmath34 s@xmath34 , density 10@xmath33 g @xmath2 , and integral scale @xmath177 km ( table 2 ) .
for these conditions , @xmath149 = 64 m ( table 3 ) and ka = @xmath178 .
the average number of flame surfaces is @xmath179 .
[ fig : andy5 ] shows that mixed regions as big as l ( and much bigger than @xmath149 ) occasionally exist .
the extent of mixing and hence the flame speed is highly variable ( fig .
[ fig : andyvofl ] ) .
most of the time , less mixing is seen than in fig .
[ fig : andy5 ] and only a few disjoint regions of burning are present , but structures like these can exist for an eddy turnover time or so .
characteristics of these regions include a carbon mass fraction much less than in the fuel , typically on the rapidly rising part of the energy generation curve in fig .
[ fig : taunuc ] , and temperatures that are a substantial fraction of the ash temperature .
the relatively high temperatures are a consequence of the temperature dependent nature of the heat capacity ( [ sec : cp ] ) .
[ sec : ledges ] the ledges seen in fig .
[ fig : andy5 ] play an important role in promoting detonation and it is thus worth spending a moment to discuss their credibility .
they have not been seen previously in any combustion simulation or experiment . for accessible terrestrial conditions
, it is believed that chemical flames in this regime would extinguish @xcite .
numerical exploration of this regime , other than with lem , has so far been impractical .
nevertheless , it has long been known that property fields in turbulence exhibit intermittent behavior , including the occurrence of transient well - mixed regions separated by cliffs ( sharp property changes ) .
this structure is seen , for instance , in 1d profile data from 3d numerical simulations @xcite .
a quantitative signature of this structure is the saturation of high - order structure - function exponents , reflecting the dominant contribution of the cliff regions to high - order intermittency statistics @xcite .
measurements by @xcite show clear evidence of the saturation of the scalar structure - function exponents .
lem structure - function exponents exhibit a somewhat slower roll - off to saturation @xcite , suggesting that the prevalence and duration of well - mixed regions in lem might be lower than actually occurs in turbulence .
this is plausible because maps in lem are statistically independent events , but the eddies that they represent actually occur in bunches because each eddy breakdown is an energy source for subsequent eddy breakdown .
this bunching contributes to the intermittency of turbulence .
there is some limited understanding of why turbulence exhibits these properties @xcite .
further clarification would provide insight into a mechanism that appears to play a key role in the timely occurrence of detonations in supernovae .
[ sec : ddt ] in order that a detonation occur , three conditions must be satisfied . first , some region must burn supersonically .
the simplest example would be a hot , isothermal volume in which the nuclear induction time scale was a constant .
as the temperature rises , the time scale decreases and for a sufficiently large volume there comes a temperature where the burning time is shorter than the sound crossing time .
the pressure will increase in that region faster than expansion can damp its growth and , after a maximum temperature is reached ( i.e. , the fuel is gone ) , the region expands at a speed that is somewhat higher than the sound speed in the surrounding medium .
in reality , the temperature is never completely isothermal , but , as we shall see , it suffices to burn only some fraction of the fuel within the volume in a sound crossing time .
a subset of fluid elements within the volume must have the same temperature to some level of tolerance that depends on their burning time scale .
the smaller the fraction , the smaller the overpressure , but sonic expansion nevertheless occurs .
the colder matter is an inert dilutant .
second , the size of the `` detonator '' must exceed some critical value @xcite .
that critical mass is larger if the mass fraction of combustible fuel is small or if the fraction of inert cold matter is large . because some mixing of ash into the pre - detonation region is unavoidable and because an appreciable amount of carbon must burn to reach temperatures where the time scales start to approach sonic , the critical masses we compute here
will be larger than those for mixtures of just carbon and oxygen with a smooth temperature structure . for a given turbulent energy ,
density must fall to lower values to obtain these larger structures .
third , and perhaps most subtly , there must exist , in a significant fraction of the mass , preferably at its edge , a nearly sonic phase velocity for the burning @xcite .
this is the `` shock wave amplification by coherent energy release '' ( swacer ) mechanism for initiating a detonation in an unconfined medium .
one requires an appreciable boundary layer where @xmath180 two properties of the flame in the sf regime assist in satisfying these three conditions .
first is the unsteady nature of the burning ( [ sec : complex ] ) . as fuel and ash
are mixed , burning may briefly almost go out , only to return with a vengeance after sufficient mixing and slow burning have occurred .
this allows the creation of regions that , after some delay comparable to a turnover time , can consume fuel at a rate faster than a single flame moving at the turbulent _ rms _ speed on the integral scale .
amplification factors of three are frequently observed , and larger values are presumably possible in rare instances .
second , as discussed in
[ sec : ledges ] , transient well - mixed structures , ledges , are a frequent occurrence in the sf regime
. turbulence does not always lead to heterogeneity on macroscopic scales .
these mixed regions have relatively constant induction time and they are at least occasionally bounded by regions in which the temperature decreases ( and fuel concentration increases ) fairly smoothly .
the characteristic size of these regions is approximately @xmath149 , which increases with decreasing density . as a result ,
the ratio of mixing time to sound crossing time decreases , since @xmath181 and this helps detonation happen .
however , as the density decreases , the critical size required to initiate a detonation also increases dramatically @xcite , so mixing a smaller region at higher density may , in some cases , be better than a larger one at low density
. detonation may occur before the characteristic size of the mixed region becomes equal to the integral scale .
another important point favoring detonation is that once burning starts to happen on a time scale approaching sonic , it no longer occurs at constant pressure .
the pressure rises with respect to the surrounding fuel and is not immediately relieved by expansion . at constant volume
, a given amount of carbon burning raises the temperature more , thus appreciably shortening the time scale for additional burning .
that is , for the relevant conditions , the heat capacity at constant pressure is appreciably larger than the heat capacity at constant volume .
the final temperature from burning all the fuel is also higher .
[ sec : itworks ] in order to verify that some of the mixed regions calculated using lem would actually detonate , the burning of select mixtures was followed using the compressible hydrodynamics code , kepler ( * ? ? ?
* ; * ? ? ?
[ fig : detonation ] ) .
the composition and temperature structure were taken from an lem simulation of flame - turbulence interaction at a density of 1.0 @xmath182 g @xmath2 for a characteristic turbulent speed of 500 km s@xmath34 on an integral scale of 10 km ( table 2 ) .
carbon had a mass fraction of 0.75 in the fuel .
use of 65536 zones gave a characteristic reynolds number of 80,000 and a resolution not achievable in a multi - dimensional simulation on this length scale .
the lem calculation was run for 0.10 s , or 5 eddy turnover times on the integral scale , during which 40 dumps were created at equal time intervals .
visual inspection isolated several cases for further study .
one of these was dump number 28 , made 70 ms after the beginning of the run .
the overall burning rate at this time was not particularly high , corresponding to an effective speed of only 250 km s@xmath34 .
however , most of the energy generation was occurring in regions where the carbon mass fraction was already low ( fig .
[ fig : detonation ] ) , so the local _ rate _ of pressure increase was quite large , even though the integrated total change in pressure in the end was small .
kepler , frequently used for studying stars and supernova explosions , is an implicit 1d hydrodynamics code with a spherical lagrangian mesh . in order to simulate a problem with approximately plane parallel geometry ,
the 4 km of interest was mapped on top of a sphere of pure ash ( mostly magnesium and silicon ) with a radius 10 km so that lem zones of constant thickness had approximately constant mass and thickness in kepler . in the figure , that 10 km has been subtracted off of the x - coordinate .
the overall thickness in km of the region of interest was the same in the kepler study as in the lem calculation .
the temperature was by no means isothermal in this region .
in fact , an isothermal runaway would not have initiated a detonation here , even if the burning were supersonic . however , several key elements favoring detonation were in place .
first , there were some regions where the temperature was already high because of previous mixing and burning .
the rate of carbon burning was quite high in these regions with a peak energy generation rate near 10@xmath183 erg g s@xmath34 .
these were embedded in an extended region with x@xmath184 @xmath185 that was already quite warm due to mixing .
finally that mixture lay at the base of a region where the carbon mass fraction increased steadily , if somewhat noisily , in an outwardly direction .
it is important to emphasize that this was the result of an lem calculation , and not an artificial construct . during the first phase , the `` initiator '' burned rapidly producing an overpressure in the surrounding zones of about 10 - 15% .
this increase was sufficient to cause the warm material to burn faster , on a roughly sonic time scale .
their expansion then compressed and ignited material at successively higher carbon mass fractions in the gradient . while weak at first , the detonation strengthened and by the last frame shown was strong enough that its permanent propagation was guaranteed .
it was followed until it left the grid . among the interesting implications here
is that a region need not be highly isothermal , nor need all of it initially burn on a supersonic time scale to provoke a detonation , though near sonic speeds are needed .
a pile up of strong acoustic pulses at around 2.3 km initiates the runaway .
[ sec : atc ] almost as interesting were the many other sample mixtures in this and other runs that , when mapped into kepler , did _ not _ detonate .
instead , the irregular burning that is characteristic of combustion in the sf regime produces strong pressure waves .
a particularly strong pulse occurs roughly every turnover time on the integral scale because these big mixing events trigger a lot of burning .
an example from the same run that produced the detonation in fig .
[ fig : detonation ] is shown about one turnover time later ( dump 39 at 95 ms ) in fig .
[ fig : pulse ] .
this time the burning did not produce a detonation , but substantial burning still occurred on less than a sonic time scale for the 4 km shown ( the sound crossing time for this region is about 1 ms ) .
this burning produced a strong pulse that sent matter moving inwards and outwards at @xmath186 km s@xmath34 , about the same value as assumed for turbulence on an integral scale of 10 km in the study . in three dimensions
, these pulses would be quasi - spherical and would occur all over the surface where burning and mixing are going on .
collisions between the fronts would pump additional energy into turbulence on a scale comparable to the width of the burning region , i.e. , the integral scale , or about 10 km .
as noted , the velocities in these pulses are frequently larger than expected from the assumed turbulent energy on that length scale .
as discussed by @xcite , this sets the stage for a potential runaway .
more turbulence leads to increased mixing which leads to more violent irregular burning , which creates more turbulence .
the culmination of this runaway could be a detonation .
g @xmath2 . for 75% carbon ,
the likely density range is 0.5 to 1.0 @xmath182 g @xmath2 .
[ fig : darm],title="fig:",scaledwidth=47.5% ] g @xmath2 . for 75% carbon ,
the likely density range is 0.5 to 1.0 @xmath182 g @xmath2 .
[ fig : darm],title="fig:",scaledwidth=47.5% ]
[ sec : concl ] three regimes of turbulent flame propagation relevant to nuclear burning in a type ia supernova have been explored . at high density , for karlovitz numbers less than about one ,
burning occurs in multiply folded laminar flames .
the overall progress of burning is governed by the turbulent energy and has a speed that is independent of the laminar speed .
a similar description of `` laminar flame brushes '' has been given many times in the literature , but this is the first time it has been simulated in a supernova ( e.g. , fig .
[ fig : flamelet ] ) .
the average number of flamelets in the flame brush is @xmath187 , and initially is not large . because of this , there will be considerable variation in the burning rate .
we estimate such variations to be as large as a factor of three .
however , detonation does not happen so early because the turbulent speed then is very subsonic and the individual flamelets are thin .
later , the number of flamelets becomes very large , reaching hundreds or even thousands , and variations in the overall burning rate are much smaller .
detonation remains impossible in the flamelet regime .
no critical mass of hot fuel can be obtained .
as the karlovitz number increases above about 10 , hot ash and cold fuel can be mixed for the first time and detonation is , in principle , possible .
two cases of turbulent burning were explored corresponding to the well - stirred reactor ( wsr ; da @xmath21 ,
[ sec : distrib ] ) and stirred flames ( sf ; da @xmath13 ; [ sec : stirred ] ) . in the wsr regime , turbulent diffusion substitutes for heat diffusion and turbulently broadened flames result .
these flames can be much larger than the integral scale . because the integral scale in a supernova is so large , this limiting case
is probably never fully realized , but is amenable to numerical simulation @xcite .
we normalized an uncertain constant in the lem calculation , @xmath70 , ( [ sec : lem ] ) to those simulations . in the stirred flame regime ,
one again has something like a flame brush , but with turbulently broadened structures substituting for the individual flamelets in the flamelet regime .
the structure of burning here is complex and highly variable . because the burning time scale is very temperature sensitive , it is possible to mix large regions of `` warm '' fuel and ash in which are embedded smaller nuggets on the verge of explosion .
mixed regions of nearly constant temperature are sometimes observed ( fig .
[ fig : andy5 ] , [ sec : ledges ] ) .
spontaneous detonation can happen in the sf regime ( fig .
[ fig : detonation ] ,
[ sec : itworks ] ) .
it does not require complete mixing on the integral scale , and hence da @xmath188 as discussed in @xcite , but is favored by values of da that are not large .
the necessary conditions occur infrequently and require a favorable confluence of several mixed regions including : a ) some region , perhaps not large and with a low mass fraction of carbon , where the burning occurs sufficiently rapidly to increase the pressure by a small fraction supersonically ; and b ) an extended region where the carbon mass fraction rises gradually and the temperature falls slowly .
neither region need be particularly homogeneous as long as some part of the composition burns supersonically for condition a ) and there are neither large barriers of ash or abrupt , sustained increases in carbon mass fraction for condition b ) . in general ,
the production of these situations require turbulent speeds that are already a considerable fraction of sonic , certainly within 10% and probably 20% .
overall the progress of burning in the sf regime varies frequently by a factor of three up and down , and rare excursions to larger values probably also occur due to intermittency @xcite .
we also find that appreciable turbulence can be put into the burning region on a scale comparable to the integral scale by small subsonic explosions of mixtures that fail to detonate ( [ sec : atc ] , fig .
[ fig : pulse ] ) . before the big bang in these stars
, there is a lot of thunder . though we have not demonstrated it here , we speculate that this energy input may exceed that put in at the large scale by the flame instabilities . if so , there is the possibility of a turbulent runaway in which mixing pumps energy into turbulence which in turn causes accelerated mixing @xcite
. the endpoint would be detonation .
these necessary conditions for detonation are summarized in fig .
[ fig : darm ] .
detonation is estimated to occur for a reasonable range of turbulent energies in the nearly horizontal band between the lines da = 10 and da = 100 .
it must be acknowledged that this figure is very approximate because the nuclear time scale ( and hence damkhler number ) are poorly determined .
the effective nuclear time scale in da is probably longer than in table 3 and intermittency essentially raises the effective turbulent speed making detonation possible at a higher density .
this is why the conditions da = 10 - 100 in fig .
[ fig : darm ] are probably more favored than say 1 to 10 . above da = 100
, the mixed regions may be too small to initiate a detonation .
three - dimensional simulations by @xcite show that the necessary degree of turbulence for detonation , roughly @xmath189 = 500 km s@xmath34 , is realized in the full star models .
we also find some dependence of the detonation conditions on initial carbon mass fraction ( see also * ? ? ?
* ) . for lower carbon mass fractions ,
damkhler numbers of order 10 are reached at a higher density .
if detonation does occur at a higher density , the explosion makes more @xmath78ni and a brighter supernova .
however , detonation depends on achieving a significant overpressure on a sonic time scale . by the time that carbon - poor fuel carbon reaches a temperature where it burns rapidly ( fig .
[ fig : taunuc ] ) , the remaining burning produces too small an overpressure .
detonating carbon - rich fuel is , in this sense , easier ( see also * ? ? ?
* ; * ? ? ?
it is important to note that the relevant location for detonation is probably in the outer layers where the density first declines below 10@xmath33 g cm@xmath34 at the flame front , not near the center .
the carbon abundance is higher in these outer layers .
all in all , our results are supportive of the hypothesis that some , perhaps even all chandrasekhar mass white dwarfs explode by a delayed detonation that occurs shortly after the burning enters the sf regime .
this would then make the location and number of detonation points , along with the ignition conditions @xcite , the principal determining factors in the intrinsic properties of a type ia supernova .
several avenues for future investigation are opened up by this work .
first , our results hinge on the calibration of the 1d lem model to direct 3d simulation @xcite .
the 3d study used for normalization here was carried out in the wsr regime , but the results ( e.g. , the constant c ) were assumed to be valid in the sf regime . given the subgrid model developed here , it should be possible to carry out equivalent 3d studies for the sf regime ( @xmath191 ) .
the results for detonation ( [ sec : itworks ] ) and active turbulent combustion ( [ sec : atc ] ) were obtained by mapping results from lem using a linear grid into a 1d compressible hydro - code with spherical coordinates .
it would be greatly preferable to see both the mixing and the strong pressure waves in the same , preferably 3d study .
because of the range in length scales , the need for a large effective reynolds number , and the rare , transient nature of the phenomena , this will require a very major investment of computational resources , but should be practical in the near future .
ultimately , of course , one would want to see these results applied to full scale models of the supernova and its light curve .
the authors gratefully acknowledges helpful conversations on the subject of the paper with andy aspden , john bell , and martin lisewski . this research has been supported by the nasa theory program ( nng05gg08 g ) and the doe scidac program ( de - fc02 - 06er41438 ) .
work at sandia was supported by the us department of energy , office of basic energy sciences , division of chemical sciences , geosciences and biosciences .
sandia is a multiprogram laboratory operated by sandia corporation , a lockheed martin company , for the us department of energy under contract de - ac04 - 94al85000 . 0.5 in
properties of the triplet map imply a simple , exact expression for the turbulent burning velocity @xmath97 in terms of the lem parameters @xmath60 ( turbulent diffusivity ) and @xmath5 ( largest allowed map size ) under the conditions @xmath192 ( here assuming the flamelet regime ) and @xmath193 ( implying re @xmath12 ) . on the lem domain ( coordinate @xmath194 ) , assume fuel on the right and ash on the left , so the flame advances rightward ( direction of increasing @xmath194 ) .
@xmath97 is evaluated by tracking the forward progress of the rightmost ( forwardmost ) ash location .
the only mechanism affecting this location , denoted @xmath195 , is the triplet map because small @xmath6 implies that the contribution of laminar burning is negligible .
any triplet map containing @xmath195 maps @xmath195 to three locations , the rightmost of which , denoted @xmath196 , exceeds @xmath195 .
specifically , if the map interval is @xmath197 $ ] , then @xmath198 , i.e. , the distance from @xmath196 to @xmath199 is one third of the pre - map distance .
the advancement of the rightmost ash location is therefore @xmath200 .
suppose that the interval @xmath197 $ ] is chosen arbitrarily .
if it contains @xmath195 , then @xmath195 is equally likely to be anywhere within the interval , and hence is uniformly distributed in the interval .
the advancement is therefore averaged over @xmath195 values in the interval .
because the advancement is linear in @xmath195 , this implies that the average value @xmath201 is substituted for @xmath195 , giving the average advancement @xmath202 , where @xmath58 is the map size .
suppose that all maps were the same size @xmath58 and denote the frequency of maps containing a given location as @xmath203 .
then the average rate @xmath97 of advancement of @xmath195 is @xmath203 times the average advancement of @xmath195 per map , giving @xmath204 .
@xmath203 can be expressed in terms of @xmath60 . from random walk theory ,
@xmath60 is one - half of the frequency of events that displace a point times the mean - square displacement per event .
the mean - square displacement induced by a size-@xmath58 triplet map is @xmath205 citepart5 , so @xmath206 , giving @xmath207 and thus @xmath208 .
this illustrates the analysis but is not the case of interest because inertial - range turbulence is represented in lem by a distribution of map sizes @xmath58 .
as explained in @xcite , in lem the map size distribution is @xmath63 for @xmath58 in the range @xmath209 $ ] , where @xmath64 is a normalization factor .
the total frequency of maps of all sizes per unit domain length is denoted @xmath68 .
the frequency of maps in the size range @xmath210 $ ] that contain a given point is then @xmath211 , so @xmath212 . for high re ,
the @xmath66 term is negligible and is dropped , giving @xmath213 . to obtain @xmath97 ,
the average advancement @xmath214 for map size @xmath58 is multiplied by @xmath215 and integrated over @xmath58 to obtain ( ignoring the @xmath66 term ) @xmath216 . in terms of @xmath60 ,
the result @xmath217 is obtained .
the numerical factor is four times larger than if all maps were of size @xmath5 .
a heuristic interpretation of this result is that the typical map size from a flame propagation viewpoint is @xmath218 when @xmath59 is based on inertial - range scaling .
cccccc 0.50 & 0.6 & 8.62(2 ) & 11.7 & 15.6 & 19.3 + 0.50 & 0.8 & 1.77(3 ) & 4.35 & 6.01 & 7.03 + 0.50 & 1.0 & 3.23(3 ) & 2.08 & 3.12 & 3.32 + 0.50 & 1.2 & 4.99(3 ) & 1.15 & 1.78 & 1.82 + 0.50 & 1.4 & 7.19(3 ) & 0.70 & 1.14 & 1.10 + 0.50 & 1.6 & 9.66(3 ) & 0.45 & 0.76 & 0.72 + 0.50 & 1.8 & 1.26(4 ) & 0.31 & 0.54 & 0.50 + 0.50 & 2.0 & 1.58(4 ) & 0.22 & 0.40 & 0.36 + 0.50 & 2.5 & 2.47(4 ) & 0.12 & 0.23 & 0.19 + 0.50 & 3.0 & 3.54(4 ) & 0.067 & 0.14 & 0.11 + 0.50 & 3.5 & 4.66(4 ) & 0.041 & 0.090 & 0.065 + & & & & & + 0.75 & 0.3 & 1.73(2 ) & 172 & 157 & 325 + 0.75 & 0.4 & 8.65(2 ) & 21.4 & 22.4 & 40.0 + 0.75 & 0.5 & 1.74(3 ) & 9.80 & 11.4 & 17.7 + 0.75 & 0.6 & 2.80(3 ) & 5.02 & 6.12 & 8.94 + 0.75 & 0.7 & 4.10(3 ) & 2.94 & 3.52 & 5.29 + 0.75 & 0.8 & 5.86(3 ) & 1.80 & 2.27 & 3.21 + 0.75 & 0.9 & 8.39(3 ) & 1.25 & 1.73 & 2.20 + 0.75 & 1.0 & 1.08(4 ) & 0.85 & 1.21 & 1.49 + 0.75 & 1.2 & 1.70(4 ) & 0.43 & 0.66 & 0.77 + 0.75 & 1.4 & 2.37(4 ) & 0.25 & 0.40 & 0.45 ccccccccc 0.50 & - & 0 & 2048 & 0.0244 & rad & 0.0323 & 2.1 & 3.1 + 0.50 & - & 0 & 256 & 0.196 & rad & 0.0314 & 2.1 & 3.1 + 0.50 & 15 & 2.47 & 2048 & 0.244 & rad . & 0.149 & 40 & 70 + 0.50 & 15 & 2.47 & 2048 & 0.244 & sg & 0.156 & 40 & 70 + 0.50 & 120 & 4.93 & 2048 & 0.977 & sg & 0.61 & 160 & 300 + 0.50 & 960 & 9.86 & 2048 & 9.77 & sg & 2.3 & @xmath24700 & @xmath241200 + 0.50 & 960 & 9.86 & 16384 & 1.22 & sg & 2.3 & @xmath24700 & @xmath241200 + 0.50 & 7680 & 19.7 & 8192 & 9.77 & sg & 8.2 & fluc .
+ 0.50 & 6.14(4 ) & 39.4 & 32768 & 9.16 & sg & 26 & fluc . & fluc .
+ 0.50 & 4.92(5 ) & 78.9 & 65536 & 49.9 & sg & 110 & fluc . & fluc .
+ 0.50 & 3.93(6 ) & 158 & 65536 & 360 & sg & 280 & fluc . & fluc .
+ 0.50 & 3.93(6 ) & 340 & 65536 & 360 & sg & 440 - 470 ?
& fluc . & fluc .
+ & & & & & & & & + 0.75 & - & 0 & 2048 & 0.0391 & rad & 0.113 & 0.85 & 1.2 + 0.75 & 10 & 2.16 & 2048 & 0.0977 & rad & 0.415 & 10 & 20 + 0.75 & 40 & 3.42 & 4096 & 0.0977 & rad & 0.922 & @xmath2425 & @xmath2460 + 0.75 & 80 & 4.31 & 8192 & 0.0977 & rad & 1.46 & @xmath2440 & @xmath2490 + 0.75 & 80 & 4.31 & 8192 & 0.0977 & sg & 1.54 & @xmath2440 & @xmath2490 + 0.75 & 320 & 6.84 & 32768 & 0.0977 & sg & 3.2 & fluc . & fluc .
+ 0.75 & 2560 & 13.7 & 32768 & 0.391 & sg & 12 & fluc . & fluc .
+ 0.75 & 2.05(4 ) & 27.4 & 32768 & 3.05 & sg & 40 & fluc . & fluc .
+ 0.75 & 1.00(4 ) & 215 & 32768 & 2.50 & sg & 96 & fluc . & fluc .
+ 0.75 & 1.00(6 ) & 500 & 32768 & 152 & sg & 1020 & fluc .
+ 0.75 & 1.00(6 ) & 500 & 65536 & 76.3 & sg & 870 & fluc .
+ cccccc 0.50 & 0.6 & 6.1(-2 ) & 4.7(5 ) & 1.0(2 ) & 3.5(3 ) + 0.50 & 0.8 & 1.2(-2 ) & 4.0(4 ) & 1.8(2 ) & 8.9(2 ) + 0.50 & 1.0 & 3.5(-3 ) & 6.4(3 ) & 3.2(2 ) & 2.5(2 ) + 0.50 & 1.2 & 1.3(-3 ) & 1.5(3 ) & 5.0(2 ) & 9.6(1 ) + 0.50 & 1.4 & 5.8(-4 ) & 4.5(2 ) & 7.2(2 ) & 4.3(1 ) + 0.50 & 1.6 & 2.9(-4 ) & 1.6(2 ) & 9.7(2 ) & 2.2(1 ) + 0.50 & 1.8 & 1.6(-4 ) & 6.4(1 ) & 1.3(3 ) & 1.3(1 ) + 0.50 & 2.0 & 9.5(-5 ) & 2.9(1 ) & 1.6(3 ) & 7.5(0 ) + 0.50 & 2.5 & 3.2(-5 ) & 5.7(0 ) & 2.5(3 ) & 2.8(0 ) + 0.50 & 3.0 & 1.3(-5 ) & 1.5(0 ) & 3.5(3 ) & 1.2(0 ) + 0.50 & 3.5 & 6.4(-6 ) & 5.1(-1 ) & 4.7(3 ) & 6.4(-1 ) + + 0.75 & 0.3 & 3.6(-1 ) & 6.8(6 ) & 1.7(1 ) & 1.8(5 ) + 0.75 & 0.4 & 6.0(-2 ) & 4.6(5 ) & 8.7(1 ) & 5.8(3 ) + 0.75 & 0.5 & 1.6(-2 ) & 6.4(4 ) & 1.7(2 ) & 1.4(3 ) + 0.75 & 0.6 & 5.3(-3 ) & 1.2(4 ) & 2.8(2 ) & 4.8(2 ) + 0.75 & 0.7 & 2.1(-3 ) & 3.0(3 ) & 4.1(2 ) & 2.1(2 ) + 0.75 & 0.8 & 1.0(-3 ) & 9.9(2 ) & 5.9(2 ) & 9.5(1 ) + 0.75 & 0.9 & 5.0(-4 ) & 3.5(2 ) & 8.4(2 ) & 4.6(1 ) + 0.75 & 1.0 & 2.8(-4 ) & 1.4(2 ) & 1.1(3 ) & 2.6(1 ) + 0.75 & 1.2 & 1.0(-4 ) & 3.3(1 ) & 1.7(3 ) & 9.4(0 ) + 0.75 & 1.4 & 4.5(-5 ) & 9.6(0 ) & 2.4(3 ) & 4.3(0 ) | the nature of carbon burning flames in type ia supernovae is explored as they interact with kolmogorov turbulence .
one - dimensional calculations using the linear eddy model of @xcite elucidate three regimes of turbulent burning . in the simplest case , large scale turbulence folds and deforms thin laminar flamelets to produce a flame brush with a total burning rate given approximately by the speed of turbulent fluctuations on the integral scale , @xmath0 .
this is the regime where the supernova explosion begins and where most of its pre - detonation burning occurs . as the density declines , turbulence starts to tear the individual flamelets , making broader structures that move faster . for a brief time , these turbulent flamelets are still narrow compared to their spacing and the concept of a flame brush moving with an overall speed of @xmath0 remains valid .
however , the typical width of the individual flamelets , which is given by the condition that their turnover time equals their burning time , continues to increase as the density declines . eventually , mixed regions almost as large as the integral scale itself are transiently formed . at that point
, a transition to detonation can occur .
the conditions for such a transition are explored numerically and it is estimated that the transition will occur for densities near @xmath1 g @xmath2 , provided the turbulent speed on the integral scale exceeds about 15% sonic . an example calculation shows the details of a detonation actually developing . |
high resolution uv spectra of @xmath10 quasars found in the sloan digital sky survey ( sdss ) provide solid evidence that the intergalactic hydrogen was in a _ completely ionized _ state up to an early epoch of @xmath0 ( e.g. , fan etal .
2001 ) . at slightly higher redshifts ,
however , a small fraction of neutral hydrogen is being seen , as indicated by the complete gunn - peterson trough found in several sdss quasars at @xmath11 6.43 ( becker etal .
2001 ; fan etal .
2003 ) . this has been used to argue that @xmath0 marks the end of the reionization era , and that the redshift at which the reionization began may not be much higher than @xmath0 ( becker etal .
2001 ; fan etal .
recently , cen ( 2003 ) proposed that the universe might have been reionized twice , where the first epoch was completed at @xmath1216 and the second epoch at @xmath0 .
the first reionization is consistent with an early beginning of reionization suggested by the recent wilkinson microwave anisotropy probe ( wmap ) results ( e.g. , spergel etal .
2003 ) , while the second reionization is consistent with the sdss results .
it is still not certain what kind of objects provided the reionizing background .
here we focus on the end of the reionization era at @xmath0 .
there seems to be a consensus that the agn population is not sufficient to account for the entire required reionizing photon budget ( e.g. , fan etal . 2002 ) , which leaves @xmath13 star - forming galaxies as the only alternative . the discovery of @xmath14 emitters at @xmath15 ( hu etal .
2002 ; kodaira etal .
2003 ) suggests that star - forming galaxies did exist at such early epochs , and the significant number of faint lyman - break galaxy ( lbg ) candidates at @xmath06.5 found by yan etal .
( 2003a ) suggest that these objects could indeed be responsible for the completion of the reionization at this epoch . in this letter
, we further study the contribution of galaxies and agn to the reionizing background , and point out that dwarf galaxies may well have contributed the vast majority of the reionizing photons .
our investigation provides significant constraints to the luminosity function ( lf ) of galaxies at @xmath0 , most importantly the faint - end slope of the lf , and the critical luminosity of the @xmath8 luminous galaxies that must have existed at @xmath0 .
this paper is organized as follows . in 2 , we summarize our current best knowledge of the lf of galaxies and agn at @xmath0 , which forms the base of our discussion .
the contribution of these two types of objects to the total ionizing background is calculated in 3 , followed by a discussion on the implications of our result in 4 .
a summary is given in 5 . throughout the paper
, we adopt a cosmology of @xmath16 , @xmath17 , and @xmath18 km@xmath19s@xmath20mpc@xmath21 .
currently , the number of confirmed galaxies at @xmath0 is still insufficient to derive their lf in a conventional way .
yan etal . ( 2002 ) presented an @xmath22 of the lf of galaxies at @xmath0 in terms of their cumulative surface density as a function of apparent magnitude .
this estimate was made by using a schechter function , @xmath23,\ ] ] with the values of its three free parameters adopted as following : the characteristic absolute magnitude @xmath24 ( measured around rest - frame 1300 ) and the faint - end slope @xmath7 obtained at @xmath3 ( steidel etal .
1996 ) were assumed to be the same at @xmath0 , and the scale factor @xmath4 was found by using the cumulative number density of @xmath25 galaxies detected in the hubble deep field north ( hdf - n ) as the normalization . in the cosmological model used in this paper , these three parameters are @xmath26 mag , @xmath27 mpc@xmath28 mag@xmath21 , and @xmath2 .
the predicted cumulative surface density based on this estimated lf is reproduced here in fig . 1 ( red curves ) with a few revisions .
all observational constraints available to date are also plotted in this figure for comparison . among them , those based on the sdss quasar hosts , the keck @xmath14 emitters and the deep acs parallel field were discussed in yan etal .
( 2002 ; 2003a / adu , and we corrected for this before applying the magnitude zeropoints that were measured at a gain of 1 @xmath29/adu
. however , this correction turns out to be unnecessary , since the hst on - the - fly pipeline already makes the extra step of dividing the gain value into the flat - fields , which is not documented in the obvious places . as a consequence
, the magnitudes of the @xmath0 candidates reported in yan etal .
( 2003a ) should be 1.5 mag brighter than their quoted values .
1 has reflected this change , which makes the acs parallel upper limit less restrictive .
other than these , the major conclusion in that paper remains unchanged .
this change does not affect this current paper . ] ) .
the three new upper limits at the bright - end are discussed in yan ( 2003b ) . in short ,
the noao-4 m mosaic lbg and @xmath14 limits are derived from a degree - sized intermediate - band survey centered on the hdf - north and south ( yan etal .
2003 , in preparation ) , and the subaru lbg limit is what we derived based on the deep subaru data of the hdf - n ( capak etal .
the data point marked as a cross is derived from the large area lyman - alpha ( lala ) survey ( rhoads & malhotra 2001 ; rhoads etal .
recently , three more measurements have become available .
stanway etal .
( 2003a , b ) used the public great observatories origins deep survey ( goods ) single - epoch data to search for @xmath0 objects .
bouwens etal .
( 2003 ) searched for such objects in their deep acs guaranteed time obsevation ( gto ) data .
most recently , the goods team released its @xmath0 candidate list derived from the stacked three - epoch acs data ( dickinson etal .
all these new observations agree with our @xmath0 lf .
the goods candidate catalog has sufficient statistics to allow us to derive not just one , but several constraints at different flux levels from 25.0 to 27.0 mag .
all these values agree very well with our lf from 23.0 to 26.5 mag . at @xmath30 mag ,
the goods counts are significantly lower than our estimate , which is possibly due to the incompleteness of the catalog at the faint - end .
to summarize , our lf broadly agrees with all known constraints , and is about the best that one can get with the available data so far . for our purpose , an analytic form of the agn lf
is preferred .
the standard " double power - law quasar lf ( e.g. , boyle , fong & shanks 1988 ) is therefore used : @xmath31 this form has been proved to be a very good representation for quasars up to at least @xmath32 .
there are four free parameters involved : the turn - over magnitude @xmath24 , the scaling factor @xmath4 , the bright - end slope @xmath33 , and the faint - end slope @xmath34 .
the sdss has discovered 6 quasars at @xmath35 in 2870 deg@xmath36 to @xmath37 mag ( see fan etal .
2003 ) . using this as the normalization , and adopting @xmath38 and @xmath39 ( fan etal .
2001 ) , we obtain @xmath40 .
the biggest uncertainty is the faint - end slope @xmath34 , which is essentially unknown for @xmath41 .
therefore three values , @xmath42 , @xmath43 and @xmath44 , are used .
the first one is the value observed at @xmath45 3 , the second one is the steepest value that we consider to be reasonable , and the last one is the limiting case ( @xmath46 ) , which is considered for illustrative purposes only .
the surface densities of agn at @xmath0 calculated based on these parameters are superposed in fig . 1 as blue curves .
we first need to answer the following question : in order to keep the intergalactic hydrogen ionized at this redshift , what should be the critical ionizing photon emission rate per unit co - moving volume ( @xmath47 ) ? we can then find out if the integrated production rate of photons at @xmath48 due to either agn or galaxies meets this critical value .
a recipe of calculating @xmath47 has been given in madau , haardt & rees ( 1999 , their eqn .
26 ; hereafter mhr99 ) : @xmath49 where @xmath50 is the ionized hydrogen clumping factor .
choosing @xmath51 , @xmath52 and @xmath53 , one finds @xmath54 . as noted by mhr99 , the time - dependent clumping factor in this expression
has be written in the form that is scaled to the value inferred at @xmath55 from the numerical simulation of gnedin & ostriker ( 1997 ) , which gave @xmath56 at @xmath55 .
the total ionizing photon production rate , @xmath57 , due to a given population of objects , can be related to its differential lf , @xmath58 , via the following form ( yan 2003b ) : @xmath59 where @xmath60 is absolute magnitude ( in the ab system ) defined around rest - frame 1300 , @xmath61 is the luminosity distance ( in mpc ) to the objects in question , and @xmath62 is the absolute magnitude of the least luminous object that should be used in the integration .
@xmath63 is a constant whose value depends on the shape of the uv - sed of these objects , which can be represented by power laws . to be specific , @xmath64 , where @xmath65 and @xmath66 are power - law indices ( in the frequency domain ) for @xmath67912 and 912@xmath68 1400 , respectively .
the constant @xmath69 is the continuum discontinuity factor across the lyman limit at 912 .
following mhr99 , we assume that the shape of the sed of a quasar has power - law indices of 1.8 at @xmath70 1050 and 0.8 at 1050@xmath712500 , respectively . using these values and the agn lf described in 2
, it is found that the contribution from agn falls far short of the critical value @xmath47 for any reasonable faint - end lf slope ( see also lehnert & bremer ( 2003 ) for their discussion at @xmath72 ) .
integrated to @xmath73 mag beyond which the existence of agn activity becomes implausible , their total contribution are @xmath74 and @xmath75 s@xmath21 mpc@xmath28 for the faint - end slopes of @xmath42 and @xmath43 , respectively . even for the limiting case where @xmath76 ,
the derived @xmath77 value is only @xmath78 s@xmath21 mpc@xmath28 .
based on 29 lbg spectra at @xmath3 , steidel et al .
( 2001 ) measured the ratio of the continuum flux at rest - frame 1300 and at the blue side of the lyman limit to be @xmath79 4.6 .
this is equivalent to taking @xmath80 in our expression for @xmath63 , and is used hereafter .
note that by using this value , we implicitly adopt the lyman continuum photon escaping fraction ( @xmath81 ) of steidel et al .
( 2001 ) , which is 1013% in absolute terms .
the index @xmath66 used in the following analysis was estimated by using a set of model spectra of galaxies with ages of 0.1 gyr ( bruzual & charlot 1993 ) , and it was found that this sed slope is close to an equivalent power - law index of @xmath82 .
we find that the original lf estimate of yan etal .
( 2002 ) , which has a relatively shallow slope of @xmath2 , can not produce a large enough @xmath57 that reaches @xmath47 .
this lf can account for about 67% of the critical value if eqn . 2
is integrated to @xmath83 mag , but the integral increases only very slowly towards fainter magnitudes and never meets @xmath47 , even when pushed to globular - cluster - type luminosities ( @xmath84 mag ) .
modifying our lf by adopting a luminosity evolution scheme of @xmath85 , such as @xmath86=3@xmath87 , will not mitigate this problem , because the value of @xmath88 is negative in a reasonable hierarchical structure formation model , and thus will make @xmath89 fainter and the @xmath57 value even less .
thus we explore other steeper @xmath7 values to see if @xmath47 can be met at a reasonable minimum luminosity .
we consider @xmath90 , and @xmath43 , where the steepest one is the critical value at which the integral of a schechter function diverges if the integration is carried to infinitely low luminosity .
for these slopes to meet the normalizing condition ( cumulative surface density of 1.37 per arcmin@xmath36 to @xmath91 mag ; see yan etal .
2002 ) , @xmath4 has to be 4.27 , 4.00 , 3.74 , and @xmath92 , respectively .
we find that @xmath47 can be reached at the following critical absolute magnitudes for the above slopes , respectively : @xmath93 8.8 , 14.6 , 15.8 , and 16.6 mag ( at @xmath0 these corresponds to apparent magnitudes of 37.9 , 32.1 , 30.9 , and 30.1 mag , respectively ) . in other words , the @xmath8 luminous galaxies that the universe should have produced by
this redshift should be at these magnitude levels or fainter in order to have a fully ionized universe at @xmath0 .
these results are shown in fig .
2 ( top panel ) .
the quantitative results above depend on three factors , namely , the clumpiness of ionized hydrogen ( @xmath50 ) , the scaling factor of the lf of galaxies ( @xmath4 ) , and the escaping fraction of lyman continuum photons ( @xmath81 ) .
the first one affects the value of @xmath47 , while the last two affect @xmath77 .
the value of @xmath47 scales linearly with the clumping factor @xmath50 . in eqn .
1 , this clumping factor is scaled to its value at @xmath55 .
we followed mhr99 and adopted @xmath52 , which is suggested by the numerical simulations of gnedin & ostriker ( 1997 ) .
should @xmath50 drop to @xmath79 20 , @xmath47 would drop by 1/3 , and a faint - end slope steeper than @xmath2 would no longer be required , since now @xmath57 would reach the critical absolute magnitude at @xmath83 mag .
similarly , steeper slopes , @xmath94 1.7 , 1.8 , 1.9 , and 2.0 would make @xmath77 reach the critical absolute magnitude at @xmath9317.0 , 17.4 , 17.7 , and 17.9 mag , respectively .
the value of @xmath57 scales linearly with the scaling factor @xmath4 of the lf . since
increasing @xmath57 has the same effect as decreasing @xmath47 , we get exactly the same answers as above , if the scaling factor increases by a factor of 1.5 . since @xmath57 also scales linearly with @xmath81 ( assuming it is not a function of luminosity ) , the same answers hold if @xmath81 increases by the same factor .
on the other hand , if either @xmath4 or @xmath81 drops by 1/3 , a faint - end slope steeper than 1.8 is necessary to reionize the universe . in this case
, @xmath47 can be reached at @xmath95 and 14.7 mag for a slope of 1.9 and 2.0 , respectively .
smaller @xmath4 or @xmath81 requires the lf extend to fainter luminosity ; however , if either of them is smaller by a factor of two , only @xmath5 can meet the reionization requirement .
these values give the critical luminosity that the least luminous galaxies should have , i.e. , a lower limit to @xmath60 . by considering the additional constraint that the intergalactic hydrogen at slightly earlier epoch still had a non - negligible neutral fraction
, we can further improve these limits into two - sided luminosity ranges .
cen ( 2003 ) gave a detailed modeling of the evolution of intergalactic medium in the double - reionization scheme , and the neutral fraction inferred from this modeling is about 15% at @xmath9 .
the precise neutral fraction at a specific redshift does not matter too much for our purpose ; the point is that at an earlier epoch , e.g. at @xmath9 , the total ionizing photon production rate of galaxies should not exceed the critical value such that a significant amount of neutral hydrogen can still exist , otherwise no gp - trough would be seen at @xmath0 .
the calculation can be performed similarly to that described in 3 . to first order
, we can assume that there is no evolution in the lf during the relatively short time period from @xmath53 to 7 , which is only 0.16 gyr .
at @xmath96 , @xmath97 s@xmath21 mpc@xmath28 for the nominal @xmath50 value of 30 .
now the requirement is @xmath98 .
if nothing else changes , a lf faint - end slope of 1.8 or shallower will satisfy this criterion .
steeper slopes of @xmath94 1.9 and 2.0 require that the lf truncates before it reaches @xmath93 12.5 and 14.6 mag , respectively .
the bottom panel of fig .
2 demonstrates these results .
if @xmath99 , or equivalently if either @xmath4 or @xmath81 is increased by a factor of 1.5 , it requires that the lf truncates before it reaches @xmath93 12.4 , 15.8 , 16.9 , and 17.5 mag for slopes of 1.7 , 1.8 , 1.9 , and 2.0 , respectively ( a cut - off is not required for a slope of 1.6 ) . on the other hand ,
if either @xmath4 or @xmath81 is decreased by 1/3 , or equivalently if @xmath100 , slopes of 1.9 and 2.0 require the lf truncate before it reaches @xmath93 4.4 and 11.8 mag .
table 1 summarizes these results .
we note that case a ) is what we believe to be the most plausible , given our best knowledge about the relevant parameters .
case b ) and c ) , on the other hand , explore a wider parameter space . however , we point out that case b ) is not very likely for several reasons .
first , the clumping factor is not likely to be as low as 20 , since the nominal value of @xmath52 obtained by gnedin & ostriker ( 1997 ) might already have underestimated the true value because of the finite resolution of their simulation .
second , the scaling factor of the lf , @xmath4 , is not likely to be significantly higher than the value suggested in yan etal .
( 2002 ) , otherwise all the recent searches for @xmath0 objects would have resulted in more bright ( @xmath101-band @xmath102 mag ) candidates than what have been actually observed .
third , the lyman photon escaping fraction of star - forming galaxies is not likely to be much higher than @xmath79 1013% , given that such a value already seems to be rather high ( cf .
giallongo etal .
2002 ) .
the ultra deep field ( udf ) campaign of the acs / hst ( http://www.stsci.edu/hst/udf ) , whose data are now being taken and will be released to the public in february 2004 , will greatly narrow down the range of @xmath7 . depending on its the real value , we predict a total of 50 to 80 genuine @xmath0 objects in the udf to its designed 10 @xmath103 depth of 28.4 mag .
in this paper we investigate the major sources of reionization at @xmath0 based on the best available observational constraints . using the lf estimate proposed for galaxies at this redshift ( yan etal .
2002 ) as the starting point , we find that @xmath13 galaxies can account for the entire reionizing background , provided that the faint - end slope of their lf is steep enough and that the lf extends to sufficiently low luminosity .
we explore a range of faint - end slopes , from @xmath2 , a value that the lf has at @xmath3 , to @xmath5 , the critical value at which the integration of a schechter lf diverges .
we find that dwarf galaxies , rather than their more luminous counterparts or agn , produced the vast majority of the reionizing photons at @xmath0 .
using the best estimate of the relevant parameters , we find that the faint - slope of the lf of galaxies at @xmath0 should be steeper than 1.6 , if the adopted @xmath4 holds , or as steep as @xmath5 if @xmath4 is lower by a factor of two . in addition , we show that the least luminous galaxies should be fainter than a certain critical luminosity , whose exact value depends on the actual faint - end slope .
furthermore , by requiring that there still was a non - negligible neutral hydrogen fraction at slightly earlier epochs ( @xmath9 ) , we point out that the lf should truncate at a certain luminosity threshold , whose value again depends on the actual faint - end slope .
the hst udf data will be able to narrow down the range of this important parameter , whose precise value is obviously important for planning the performance of the james webb space telescope .
funding from nasa / jwst grant nag5 - 12460 is acknowledged .
the authors would like to thank dr .
mark dickinson for pointing out the hidden step in the acs pipeline reduction that led to the improper zeropoint used in yan etal .
( 2003a ) .
we also thank drs .
piero madau and joe silk for helpful discussions .
we thank the referee for the helpful comments .
becker , r .
h. , et al .
2001 , aj , 122 , 2850 bouwens , r. j. , et al .
2003 , apj , 595 , 589 boyle , b. j. , fong , r. , & shanks , t. 1988 , mnras , 231 , 897 bruzual , a. g. , & charlot , s. 1993 , apj , 405 , 538 cen , r. 2003 , apj , 591 , 12 dickinson , m. , et al . 2003 ( astro - ph/0309070 ) fan , x. , et al .
2001 , aj , 122 , 2833 fan , x. , et al .
2002 , aj , 123 , 1247 fan , x. , et al .
2003 , aj , 125 , 1649 gnedin , n. y. , & ostriker , j. p. 1997 ,
apj , 486 , 581 giallongo , e. , cristiani , s. , dodorico , s. , & fontana , a. , 2002 , apj , 568 , l9 hu , e. m. , et al .
2002 , apj , 568 , l75 kodaira , k. , et al .
2003 , pasj , 55 , l17 lehnet , m. d. & bremer , m 2003 , apj , 593 , 630 madau , p. , haardt , f. , & rees , m. j. 1999 , apj , 514 , 648 rhoads , j. e. , et al .
2003 , aj , 125 , 1006 rhoads , j. e. & malhotra , s. 2001 , apjl , 563 , 5 spergel , d. n. , et al .
2003 , apjs , 148 , 175 stanway , e. r. , bunker , a. j. , & mcmahon , r. g. , 2003a , mnras , 342 , 439 stanway , e. r. , bunker , a. j. , mcmahon , r. g. , ellis , r. s. , treu , t. , & mccarthy , p. j. , 2003b ( astro - ph/0308124 ) steidel , c. c. , pettini , m. , & adelberger , k. l. 2001 , apj , 546 , 665 yan , h. , windhorst , r. a. , odewahn , s. c. , cohen , s. h. , rttgering , h. j. a. , & keel , w. c. 2002 , apj , 580 , 725 yan , h. , windhorst , r. a. , & cohen , s. h. 2003a , apj , 585 , l93 yan , h. 2003b , ph .
d. dissertation , arizona state university @xmath7 & @xmath104 & @xmath105 & @xmath105 & @xmath105 + 1.6 & 4.55 & n / a & @xmath106 & n / a + 1.7 & 4.27 & @xmath107 & ( 17.0 , 9.1 ) & n / a + 1.8 & 4.00 & @xmath108 & ( 17.4 , 14.6 ) & n / a + 1.9 & 3.74 & ( 15.8 , 12.6 ) & ( 17.7 , 15.9 ) & ( 12.5 , 4.4 ) + 2.0 & 3.49 & ( 16.6 , 14.7 ) & ( 17.9 , 16.6 ) & ( 14.7 , 11.8 ) + | in this paper , we address which sources contributed most of the reionizing photons .
our argument assumes that the reionization ended around @xmath0 and that it was a relatively quick process , i.e. , that there was a non - negligible fraction of neutral hydrogen in the universe at somewhat earlier epochs . starting from our earlier estimate of the luminosity function ( lf ) of galaxies at @xmath0
, we quantitatively show that the major sources of reionization are most likely galaxies with @xmath1 .
our approach allows us to put stronger constraints to the lf of galaxies at @xmath0 . to have the universe completely ionized at this redshift
, the faint - end slope of the lf should be steeper than @xmath2 , which is the value measured at lower redshifts ( @xmath3 ) , unless either the normalization ( @xmath4 ) of the lf or the clumping factor of the ionized hydrogen has been significantly underestimated . if @xmath4 is actually lower than what we assumed by a factor of two , a steep slope close to @xmath5 is required .
our lf predicts a total of 50 80 @xmath0 galaxies in the hst ultra deep field ( udf ) to a depth of @xmath6 mag , which can be used to constraint both @xmath4 and @xmath7 .
we conclude that the @xmath8 luminous galaxies existing at this redshift should reach as low as some critical luminosity in order to accumulate the entire reionizing photon budget . on the other hand ,
the existence of significant amounts of neutral hydrogen at slightly earlier epochs , e.g. @xmath9 , requires that the @xmath8 luminous galaxies should not be fainter than another critical value ( i.e. , the lf should cut - off at this point ) . |
the analysis of the klm collaboration data @xcite for pb(158 gev / nucleon)-ag / br collisions indicated the existence of events with `` strong dynamical fluctuations '' . the use of factorial moments as a tool for preliminary selection of events was suggested . in our previous paper @xcite we analyzed the cluster structure of events resulting from the heavy ion collision data and from the monte carlo ( mc ) generators .
we investigated the dependence of the cluster distributions on the parameter defining a cluster size .
the quantity we worked with was the second scaled factorial moment as a function of averaged multiplicity @xmath0 with `` two subtracted '' ( i.e. @xmath0 - 2 ) : @xmath1 we gave also a motivation for the choice of such a discrimination tool . in full agreement with the suggestions following from ref .
@xcite we have found two different patterns in the first four published experimental events . in this note
we would like to present some further analysis concerning the same subject .
we again employ the cluster definition used in the procedure implementing the bose - einstein ( be ) effect in mc @xcite and in what follows only clusters containing at least two hadrons are taken into account .
as before , we work in a two - dimensional momentum space ( @xmath2 ) since in the data set @xcite only the pseudorapidity @xmath3 and the azimuthal angle @xmath4 are given .
as a distance measure for clustering we use @xmath5 which provides also rather stable results under some coefficient change in front of these two terms defining @xmath6 . in our procedure originally each particle is considered as a single cluster .
we fix the value of a `` cluster size parameter '' @xmath7 ( the upper limit of @xmath6 , for which two particles are joined into one cluster ) and perform the clustering procedure for all pairs of particles .
our procedure is different from the clustering procedure applied in @xcite .
details are given in @xcite .
here we would only like to point out that our results are significant if we have a sufficiently large number of clusters ( with at least two particles ) .
this limits the possible range of @xmath8 values , which must neither be too small nor too large .
twenty values of @xmath8 ( @xmath9 ) were taken for each of the events under consideration .
thus for each value of @xmath8 for a given event we have * a number of particles in each cluster @xmath10 * a number of clusters with @xmath11 , @xmath12 . obviously in the following @xmath13 . with this information
we may also calculate for subsequent values of @xmath8 the entropy @xcite @xmath14 where the summation runs over all clusters in an event ( leaving aside one particle clusters ) .
the quantity @xmath15 is a probability to find a particle in the k - th cluster .
as well known , the jaynes - shannon entropy introduced in this way is a good measure of the `` amount of uncertainty '' represented by a discrete probability distribution
. therefore one may expect that it will help ( in a natural way ) to distinguish between different heavy ion collision events .
here we do not use in our clustering procedure the principle of event entropy maximisation .
a futher study of this interesting subject is in progress .
we performed the clustering procedure for * the `` random events '' of similar multiplicities obtained by using a plain uniform random generator of @xmath16 points ( serene events ) * the four klm events presented in @xcite * the events obtained from the venus generator @xcite . we have chosen randomly for presentation 5 out of 20 events generated for both the serene and venus classes .
the results for cluster distribution are presented in figs .
1a)-3a ) .
the number of clusters for all the serene events has a rather broad maximum for @xmath17 between 5 and 10 ( @xmath18 ) with the value around 200 . for venus events the maximum occurs at smaller @xmath17 and
the subsequent decrease is much faster .
two of the experimental events resemble venus events , the other two are more similar to serene events . the striking structure seen in figs .
1a)-3a ) appears again if we now plot ( fot the same range of @xmath8 ) the entropy defined in eq .
( 3 ) , as shown in figs .
1b)-3b ) .
the maximum for all the serene events is very flat , whereas for all the venus events the values of entropy decrease rather fast for higher values of @xmath8 .
the experimental events are again much more differentiated .
more quantitative comparison of the events is possible if we define for each event * the ratio @xmath19 of the maximal number of clusters in each event to the number of clusters for the last bin in @xmath8 in this event ( cf . figs .
1a)-3a ) ) * the ratio @xmath20 of the maximal value of the entropy in each event to its `` last bin '' value ( cf
1b)-3b ) ) .
the values of these ratios are given in tab.1 .
we checked that omitting clusters with only two particles we get similar results as shown in tab.2 , although the number of clusters and the values of entropy ( especially for small values of @xmath8 ) are strongly reduced .
the values of considered ratios for all serene events are very similar . for venus events the values are much bigger and
more differentiated .
for two experimental events we find the values similar to those of serene events and for two others the values are more similar to venus events . [ cols="^,^ " , ]
comparing the results presented in fig.1 and fig.2 and the numbers given in tab.1 and tab.2 we may conclude that : * all the serene events follow one ( `` quiet '' ) pattern , where the fall of the number of clusters and entropy from its maximal value is rather slow ; the broad entropy distribution reflects the randomness of these events structure * on the contrary , all the venus events exhibit a completely different behaviour ; the majority of particles goes into one or a few clusters for larger values of @xmath8 , which results in a fast fall of the number of clusters and entropy * two of the experimental events ( the same as in ref .
@xcite and ref .
@xcite ) are serene like , the other two resemble more the venus like pattern .
thus , using a slightly different language , we confirmed our previous results @xcite .
the events from data seem to be more differentiated than those from the generators .
data with full momentum measurements and particle identification would be of great help in further analysis , improving our understanding of multiple production in heavy ion collisions .
we used the data kindly provided to us by b. wosiek ( ref .
@xcite ) and the data prepared from the `` venus '' generator .
an interesting conversation with p. bogu concerning the application of the j - s entropy in medical data bases should be noticed . the financial support of kbn grants # 2 p03b 086 14 and # 2p03b 010 15 is gratefully acknowledged .
one of us ( rw ) is grateful for a partial financial support by the kbn grant # 2 p03b 019 17 .
k. fiakowski and r. wit , _ acta phys .
polon . _ * 30 * , 2759 ( 1999 ) .
k. fiakowski , r. wit and j. wosiek , _ phys .
rev . _ * d58 * , 094013 ( 1998 ) .
jaynes , _ phys
. rev . _ * 106 * , 620 ( 1957 ) .
k. werner , _ phys .
rep . _ * 232 * , 87 ( 1995 ) . + | we analyse the cluster structure of the final multihadron states resulting from heavy ion collisions using the concept of jaynes - shannon entropy .
further evidence for an interesting differentiation of events is provided . 23.0 cm 16.0 cm -2.0 cm -1.0 cm -1.0 cm k. fiakowski , r. wit _ m. smoluchowski institute of physics + jagellonian university + 30 - 059 krakw , ul.reymonta 4 , poland _ pacs : 25.70.-z + _ keywords : _ heavy ion collisions , cluster analysis , monte carlo + 4 october , 1999 + |
theories with extra dimensions have long played a role in attempts to unify other forces with gravity @xcite .
traditional ideas about hiding extra dimensions involved making them compact and small @xcite ( generally assumed to be of the order of the planck length @xcite ) , so that propagation of standard model matter in the extra dimensions requires energy of the inverse of the size of the extra dimensions .
thus , if the extra dimensions are small enough , they effectively decouple from the low - energy theory .
the mechanism of confining standard - model fields on @xmath1-dimensional subspaces ( 3-branes , or just branes ) @xcite of a higher - dimensional manifold leads to the possibility of scenarios with large extra dimensions . in models where the spacetime geometry is of a simple factorizable form
, the space of extra dimensions ( the bulk ) may be compact and perhaps as large as a millimeter @xcite .
if the spacetime geometry has a nonfactorizable form , the extra dimensions may be warped and noncompact as in the work of randall and sundrum @xcite .
the possibility of warped noncompact extra dimensions has extended our intuition about how extra spatial dimensions are manifest in four - dimensional effective field theories by showing that even if gravity propagates in non - compact higher dimensional spaces , four - dimensional observers may still empirically deduce a four - dimensional newton s law .
there has been a great deal of recent activity studying various aspects of cosmology in large extra dimension scenarios .
nonetheless , model building is still in its infancy and general features are still being uncovered . as a contribution to this effort
, we examine here whether these large extra dimension scenarios possess an analog of the `` flatness problem '' existing in four - dimensional friedmann robertson walker ( frw ) cosmology .
we find that there is most likely a higher dimensional flatness problem of character significantly different from that of the frw flatness problem .
furthermore , unlike the flatness problem in an frw cosmology , we will argue this problem is not easily solvable by inflation .
we do not present our analysis in the context of any concrete realistic model . as there is no unified theory that can address the question of initial conditions , our conclusions necessarily
must be based on certain ( plausible ) assumptions .
it is impossible to know if the fundamental theory will somehow naturally circumvent the difficulty we discuss .
also , we only address the issue of large ( even infinitely extending ) extra dimensions . we know that if the extra dimensions are macroscopic , effective field theories will be valid to describe the spacetime behavior and the calculations should be reliable . in the rest of the introduction
, we shall lay out the assumptions under which our arguments apply .
however , we first discuss the frw flatness problem in a way that most closely parallels the flatness problem of large extra dimensions .
one way of viewing the frw flatness problem is as a fine - tuning problem , associated with the fact that if the equation of state obeys @xmath2 , _ any _ deviation from spatial flatness in the early universe would , in a few expansion times , cause the universe either to collapse or to expand and become negative curvature dominated
. even an initial condition of exact flatness is problematic since spatially flat patches much larger than planck size are unnatural because the planck scale sets the energy scale associated with quantum gravitational fluctuations early in the universe .
that is , even if the universe were initially globally frw and perfectly spatially flat , any fluctuations would have destabilized spatial flatness early on , especially when the curvature of the spacetime was large enough for quantum gravitational fluctuations to be large .
note that even without introducing the issue of quantum gravitational fluctuations there is the question of why the universe chose zero spatial curvature over a region that is much larger than the planck scale to such a large degree of accuracy .
there seem to be two options : ( i ) the frw universe was nucleated by some quantum cosmological process to have exactly zero spatial curvature , or ( ii ) the frw universe arose from a planck - size flat patch which inflated to become our universe .
the first of these options is possible , but not under computational control .
the second of these options , usually referred to as inflation , is under computational control .
the attraction of inflation , besides computational control , is that a physical mechanism to generate a large flat patch out of a tiny flat patch simultaneously solves other cosmological fine - tuning problems .
it should be noted that even if inflation did not solve these other cosmological fine - tuning problems , the solution to the flatness problem alone would be enough reason to consider it seriously since there is no other known physical mechanism that can solve the flatness problem .
one may understand the frw flatness problem as due to the fact that there is no symmetry that prefers flat spatial geometries , and even if the spatial geometry was spatially flat , there is no symmetry to protect the flatness .
if one states the frw flatness problem in those terms , it is easy to appreciate the problem we discuss .
we will point out that lorentz invariance on our brane requires that the _ entire _ spacetime possess an approximate @xmath0 isometry ( not just our brane ) .
since bulk curvature would lead to @xmath0 violation observed on our brane , in the absence of some symmetry there must be some cosmological mechanism to flatten the bulk .
the relation between bulk flatness and lorentz invariance is easier to appreciate if one notes that even in four - dimensional frw models , the reason we observe lorentz invariance today can be connected to the fact that our universe is flat and old .
one begins to understand this by noting that the equivalence principle of general relativity , which protects lorentz invariance of the ultraviolet ( uv ) limit of a field theory living on a smooth lorentzian manifold , does not protect _ infrared ( ir ) _ physics from obtaining what appears to be lorentz violating terms .
this statement is in some sense obvious from the fact that field solutions generically may break the symmetry of the underlying spacetime ( a type of spontaneous symmetry breaking ) .
in fact , the curvature of spacetime generated in a four - dimensional frw universe breaks the lorentz isometry of the zero - energy vacuum spacetime .
let us see this explicitly .
if the frw metric is written as usual as @xmath3 we can write the action of a free scalar field in this background as @xmath4 \left[(\partial _ { 0}\phi ) ^{2 } - ( 1 - 2h_{0}\delta t+ ... ) ( \partial _ { i}\phi ) ^{2 } \right ] , \ ] ] where @xmath5 is the present ( @xmath6 ) expansion rate and @xmath7 .
the different coefficients of @xmath8 and @xmath9 signal lorentz violation . clearly if the ir cutoff for the field theory is taken to be @xmath10 , then @xmath11 , and one can work with a lorentz invariant field theory @xcite .
in other words , the field theory on the tangent space of the manifold is lorentz invariant only if the ir cutoff is taken to be larger than @xmath12 .
of course in this four - dimensional frw case this field theory is really only valid on large scales ( small uv energy cutoff ) , since the frw metric is only valid on large scales during most of the history of the universe .
the only reason we ever deduced lorentz invariance in the first place is because the spacetime curvature associated with the energy density in our universe is much smaller than the energy scales associated with our physics experiments involving the standard model of particle physics : we just ca nt probe energy scales smaller than @xmath12 ( a similar argument applies to planetary curvature , etc . ) .
the reason energy scales associated with our physics experiments are small compared to the frw spacetime curvature scale ( the hubble expansion rate @xmath12 ) is because the universe is old , i.e. , the flatness problem @xcite . since inflation solves the flatness problem , we can then say that our ability to observe lorentz invariance today has much to do with the fact that inflation made our universe flat ( neglecting any unaddressable metaphysical issues such as the anthropic principle ) .
extending the analogy to extra dimension scenarios is not straightforward , however , because we do not really have a standard time - dependent model of brane / bulk cosmology including its `` birth '' @xcite .
what we first establish in the next section of this paper is that departures from the @xmath0 isometry for the _ large scale _ higher dimensional background geometry will result in lorentz violations for _ any _ field living in the bulk @xcite .
we will find that the nature of lorentz violations from bulk curvature is significantly different from the nature of lorentz violations in a four - dimensional frw universe .
namely , we will see that in some cases , the lorentz symmetry is never recovered as the ir cutoff is taken to be arbitrarily large , unlike in the 4-d frw case .
this means that the four - dimensional equivalence principle does not necessarily result from a higher dimensional equivalence principle .
this is what we will call the `` inequivalence principle . ''
hence , if we assume that the observed four - dimensional gravitational theory is lorentz invariant ( obeys the equivalence principle ) to a high accuracy , say to scales of order @xmath13 , an explanation is required . as in the 4-d frw model ,
two general classes of explanations exist .
one is that for some reason the quantum gravitational dynamics is driven to initiate an @xmath0 invariant universe , and the other is that a field theoretical dynamical mechanism exists to drive the system to approximate @xmath0 invariance .
the first explanation can not be meaningfully addressed because we do not know the fundamental unified theory .
indeed , if such a mechanism existed , then that will most likely allow an alternative to inflation for solving the flatness problem .
we thus take the latter approach , and make the assumption that the universe initially was in an @xmath0 isometry violating state , which is generic since there is no symmetry to protect @xmath0 isometry .
( for instance , @xmath14 susy is compatible only with @xmath15 and @xmath0 isometries , but is generically broken with the presence of matter energy density . )
thus we will require spacetime expansion , which we will call bulk inflation , to eliminate the @xmath0 violating terms .
we find that in the most straightforward inflationary scenarios , the warp factor does not survive bulk inflation .
furthermore , we point out that for the large , flat , compact extra dimension scenarios , inflation at the effective field theory approximation can not be used to flatten out any significant @xmath16 violating curvature because that would force the initial compact dimension size to be too small , i.e. , smaller than the fundamental planck length .
this latter point has been addressed to a certain extent in ref .
finally , we find that @xmath0 violating curvature terms in the bulk can be experimentally constrained .
the order of presentation will be as follows .
we first discuss how apparent lorentz violations arise in effective field theories in a curved spacetime in higher dimensions .
we then discuss in detail why inflation generically would be necessary in noncompact warped extra dimension scenarios .
finally , we discuss the possibility of finding evidence for extra dimensions through experimental observation of lorentz violations and conclude .
it is well known that in four dimensions , the ir limit of any effective field theory is sensitive to the background spacetime curvature , which generically induces lorentz violations .
due to the equivalence principle , this lorentz violation effect generically vanishes in the limit that the ir cutoff is taken to be large .
however , what is not as well known ( at least it is new to us ) is that for any single dimensionally reduced field , increasing the ir cutoff will not lead to the recovery of lorentz invariance if the @xmath0 violating curvature is due to warping in the extra dimensions .
hence , @xmath0 violating curvatures from extra dimensions induce a four - dimensional theory that violates the principle of equivalence .
this is what we have referred to as the inequivalence principle .
another way of stating the inequivalence principle is that the equivalence principle of a higher dimensional gravitational field theory does not necessarily guarantee that the equivalence principle will be manifest for the dimensionally reduced effective field theory .
a perhaps noteworthy observation is that it is only an accident that we , as four - dimensional low energy observers , discovered lorentz invariance as a fundamental symmetry of nature .
if the total number spacetime dimensions is four and if we had been unlucky , spacetime curvature would have prevented us from ever deducing lorentz invariance until our experiments reached an energy level above the curvature scale .
as we will detail below , the situation in the brane scenario could have been worse .
the inequivalence principle tells us that if we had been unlucky to be embedded in a higher dimensional spacetime with no @xmath0 isometry , it would have been very difficult to deduce that gravitons obey lorentz invariance , _ even at short distances _ ( short compared to the background curvature scale )
. this would be true irrespective of the energy probed by experiments , even if the energy were higher than the background curvature energy scale .
this loss of lorentz invariance in the uv limit is a signature of extra dimensions ; it can not be reproduced with any purely four - dimensional background curvature .
the loss of lorentz invariance is connected to dimensional reduction . to see this ,
consider a higher dimensional spacetime .
finding a four - dimensional effective field theory description of a field propagating in higher dimensions usually requires introducing an infinite number of four - dimensional fields ( a `` kk tower '' ) and integrating over the coordinates of the extra dimensions .
let us denote these four - dimensional fields by @xmath17 .
the procedure for obtaining a four - dimensional effective field theory preserves the isometries , and thus the effective field theory for the @xmath17 is lorentz invariant if the underlying higher dimensional spacetime possesses an @xmath18 isometry with respect to the four - dimensional coordinates of interest .
now , what about the condition on the ir cutoff ?
an interesting feature of the four - dimensional effective theory for @xmath17 is that none of the _ individual _ kinetic terms for @xmath17 will recover lorentz invariance even if one takes the ir cutoff to be large .
this is because the lorentz invariant tangent space of the higher dimensional manifold is never four dimensional ! in other words , the ir cutoff scale for any single @xmath17 ( for fixed @xmath19 ) is constrained to be be less than @xmath20 where @xmath21 is the length scale of the extra dimensions , because @xmath20 is the uv cutoff scale for @xmath17 ( the field theory has been coarse - grained over length scales of @xmath20 ) .
one can describe the propagation in a five - dimensional manifold up to resolution of @xmath22 only when all the fields up to mass @xmath23 are taken together .
however , each individual @xmath17 never recovers lorentz invariance .
hence , we deduce two useful sufficient conditions for the existence of lorentz violation in the four - dimensional effective theory of a field propagating in higher dimensions . first , there must be no @xmath0 isometry in @xmath24 dimensions , and second , the coarse - graining length scale ( effective ir cutoff ) must be much larger than the radius of curvature @xmath25 : i.e. , @xmath26 .
we now present toy models to illustrate the nature of lorentz violating terms induced from extra dimensions .
consider a metric of the form @xmath27 where @xmath28 are coordinates of the bulk and the @xmath29 dependence of @xmath30 and @xmath31 makes the global geometry nonfactorizable as in the model of ref .
since we are interested in cosmological solutions we have assumed that @xmath30 and @xmath31 are independent of the spatial coordinates of the brane and only depend on @xmath32 , the comoving time coordinate of our universe .
furthermore , consider the situation in which we are interested in physics for energy scales much larger than @xmath33 and @xmath34 .
then in the adiabatic approximation we can neglect the time dependence of @xmath30 and @xmath31 and set the time variable to a particular value @xmath35 . under these conditions ,
our generic manifold is approximately described by @xmath36 where in general , @xmath37 because there is no fundamental symmetry imposing ( or protecting ) the condition @xmath38 .
this implies that there is no @xmath0 isometry generated by the lorentz group acting on the coordinates @xmath39 .
although the randall - sundrum metric containing a single extra dimension with coordinate @xmath40 ( @xmath41 ) has an @xmath0 isometry ( @xmath42 ) , the question to address is why did the metric evolve to this form .
brane solutions in which the bulk is not @xmath0 isometric were recently constructed in ref .
@xcite , where the metric is given by @xmath43 with @xmath44 , where @xmath45 is a cosmological constant in the bulk and @xmath46 is a free constant parameter . if @xmath47 , then this metric breaks the @xmath0 isometry @xcite .
now consider a simple toy model that illustrates the breaking of the @xmath0 isometry .
assume we live in five dimensions and the extra spatial dimension has the topology of @xmath48 with the radius of @xmath49 equal to @xmath50 .
suppose the background stress energy is arranged to give the line element @xmath51 where again , @xmath40 is the extra dimension coordinate @xcite .
note that this metric is not diffeomorphic to that written by randall and sundrum @xcite ; notably , it is not a conformally flat spacetime .
it is a static version of the cosmological spacetime presented in ref.@xcite .
now suppose that a free bulk scalar field @xmath52 lives in the background of this spacetime .
we will assume that its contribution to the vacuum energy has been fine - tuned to zero ( i.e. , this is not the bulk field determining the background geometry of the spacetime ) , and consider what this field looks like to a four - dimensional observer living on our brane .
the action for this bulk scalar field is @xmath53 where @xmath54 is the @xmath55-dimensional metric of eq.([metric ] ) .
we can consider what a four - dimensional observer will see by expanding this scalar field in a particular orthogonal basis @xmath56 where the @xmath57 satisfies @xmath58=-m_{n}^{2}h_{n},\ ] ] and the self - adjoint derivative condition implies @xmath59 the basis functions @xmath60 satisfying these conditions can be written down explicitly : @xmath61.\ ] ] if we insist on the normalization @xmath62 we have the normalization constant @xmath63 putting this expansion into eq .
( [ 5dscalar ] ) , we find the effective action ( @xmath64 ) @xmath65 , \label{dimreduced}\end{aligned}\ ] ] where we have defined an infinite dimensional matrix @xmath66 which characterizes the lorentz noncovariant structure of the theory .
note that for a fixed index @xmath67 and a given field @xmath68 , only the off - diagonal components of @xmath69 seem to be responsible for the lorentz non - covariant structure , because for the diagonal component we can always rescale the coordinates to obtain the usual covariant form .
however , because the standard model fields confined to the brane reveals the `` true geometry '' of the underlying spacetime , it is not true that the field redefinition completely hides the apparent lorentz violation even for a given diagonal sector .
for example , if @xmath70 we can take @xmath71 in the first line of eq . ( [ dimreduced ] ) to obtain @xmath72 apparently recovering lorentz invariance for what we will call the @xmath73 sector .
it is manifest that two points that are a distance @xmath74 apart ( as measured by the standard model physics of fields confined to our brane ) are seen to be only a distance @xmath75 apart from the point of view of the effective four - dimensional field , which actually lives in higher dimensions .
of course there is no global coordinate transformation that one can make to have all diagonal @xmath76 sectors lorentz covariant simultaneously .
hence , at least with the off - diagonal terms neglected , each field @xmath77 lives in a different apparent geometry ; i.e. , the effective distance that each field sees through its propagator is different even though the `` true '' distance in spacetime as measured by the standard model fields confined to the brane is the same .
the distance @xmath75 is what an observer would deduce from an `` inverse - square law '' analysis , and hence we will refer to it as the `` inverse - square distance . ''
the ratio of the inverse - square distance to the distance measured by fields confined to the brane is in this case just @xmath78 .
the fact that the ratio is not unity is nothing more than a consequence of the fact that the lightcone in the extra - dimensional spacetime is different from the lightcone of a field confined to the brane , as was discussed in ref .
in other words , causal signals can take a shortcut through the extra dimensions to get to a point on the brane that is farther than where a causal signal confined to the brane can go for a fixed time . from a @xmath1-dimensional point of view
, the higher dimensional signals seem acausal .
it is important to note that this noncovariant structure is independent of the basis chosen , and there is no coordinate redefinition nor field redefinition that will truly restore the lorentz symmetry . even more importantly , since the lorentz violation structure is governed by the quantity @xmath79 ( which is independent of @xmath80 , the ir cutoff of the @xmath81 dimensional momenta ) , increasing @xmath80 does not lead to the recovery of lorentz invariance for any one field @xmath82 .
hence , the inequivalence principle is manifest . as argued before , the fact that @xmath83 is a result of the fact that the underlying higher dimensional spacetime does not possess an @xmath0 isometry .
mathematically , this merely amounts to the fact that the partial differential equations governing the modes are not separable in the chosen coordinate directions .
let us now examine the magnitude of these effects .
the magnitude of the lorentz - violating effects is characterized by the coefficients @xmath84 . first note that in this model the lorentz violating effects associated with the zero mode are not very large , because @xmath85 . in particular ,
the distances are only scaled by @xmath86 . however , since the scalar field propagators behave approximately as the graviton propagator for newton s law , one can see that the `` inverse - square '' distance vs. luminosity distance comparisons can discriminate such scalings @xcite .
we leave a more careful analysis of the observables to a future study .
however , in the last section , we will explicitly show one possible experimental observable which is within the reach of upcoming gravitational experiments .
the effects for the higher mass mode can be extremely large , even if @xmath87 is much larger than electroweak energy scale .
for instance , if @xmath88 , we have @xmath89 which will be much larger than unity .
moreover , it is not always possible to treat the mixing of the zero - mode mass eigenstates with massive kk mass eigenstates as a perturbation because the mixing with massive modes can be equally large if @xmath90 is large , as can be seen by @xmath91 in general , the zero mode truncation of the effective field theory in the @xmath92 limit is not valid because @xmath93 is much greater than unity if @xmath79 is much greater than unity , and the field theory must be considered from a higher dimensional point of view .
this may be true even though the nonzero modes ( the zero mode is massless ) can be quite massive since @xmath87 may need to be large enough to hide the higher dimensional behavior of gravity .
explicitly , the mass spectrum of the nonzero modes is given by @xmath94 which would naively justify decoupling if @xmath95 or @xmath87 were sufficiently large .
however , here in general , the mass eigenstates will not be momentum eigenstates , and there does not seem to be decoupling .
what is clear , however , is that if we insist on a four - dimensional point of view , we have a theory in which the field labeled by different masses see a different effective geometry , i.e. , the inverse - square distances corresponding to the same spacetime geometry distance are different for different four - dimensional effective fields .
although the exact nature of the lorentz violating effects characterized by @xmath93 is model dependent , its magnitude can be read off from the metric of the form given in eq.([metricgeneral ] ) .
it is easy to show that in general whether @xmath93 is greater than or less than unity is roughly governed by the ratio @xmath96 where @xmath97 denotes an average over the extra dimensions .
hence , for spacetimes with @xmath98 we have an `` acausal '' effective theory while for @xmath99 , we have merely the momentum nonconserving lorentz violating effects with respect to the off - diagonal terms of @xmath93 . note
that the existence of lorentz violation generalizes to higher spin fields .
consider the graviton field @xmath100 , which is defined to be the zero mode of the metric perturbation @xmath101 + g_{mn}du^{m}du^{n } , \ ] ] where @xmath100 only depends on @xmath1-dimensional coordinates .
the graviton kinetic term will contain a term @xmath102 , \label{eq : gravitonkinetic}\ ] ] which is again lorentz violating when integrated over the extra dimension coordinates @xmath29 .
again , the metric implied by measurements of the brane - confined fields will be different from the constant metric obtained after integrating over @xmath29 in eq.([eq : gravitonkinetic ] ) .
since we have little control over the effective field theory in the context in which the extra dimensions are compactified to be planck size , we will not discuss that scenario here .
however , in the case in which the extra dimensions are large and flat @xcite or in the case in which the extra dimensions are warped and noncompact @xcite , we can ask in the context of an effective field theory whether there may be a flatness problem as outlined in the introduction . as we shall argue , in these large extra dimension scenarios there are additional flatness problem complications that did not arise in the four - dimensional frw cosmology . in
what follows , we shall first identify the flatness problem in the warped extra dimensions scenario and then discuss the case of the large extra dimension scenario . as discussed more fully in the previous section , a higher dimensional spacetime described by eqs .
( [ eq : genmetricadiabatic ] ) and ( [ eq : lorentzviolate ] ) implies that any dimensionally reduced effective theory , including gravity , will generically violate lorentz invariance . in the view of treating gravity as a theory of vielbeins in minkowski spacetime ,
this means that the vev of the vielbeins spontaneously break lorentz invariance .
hence , in warped extra dimensions scenarios , it is crucial to explain why there is an @xmath0 isometry in the extra dimensions to an accuracy that allows four - dimensional gravity ( or any other dimensionally reduced field ) to be lorentz invariant @xcite .
hence , in analogy to the frw flatness problem , if we assume that the bulk fields dimensionally reduced to four dimensions are observed to obey lorentz invariance ( as the ir cutoff is taken to be arbitrarily large ) , we have an extra - dimensional flatness problem . this analogy is summarized in table 1 . .[sigmas ] the analogy between the frw flatness problem and the extra - dimension flatness problem . [ cols="^,^",options="header " , ] if a typical initial condition of early cosmology contains @xmath0 isometry violating curvature , the extra dimensional flatness problem is real and inflation may be required to eliminate it .
note that here we are making a crucial assumption that there is no fundamental reason ( such as fundamental symmetry arguments or dynamical arguments ) to choose an @xmath0 isometric manifold as the initial condition .
indeed , if there were such a mechanism , then one may be able to modify it and utilize it to replace inflation altogether .
given that we use inflation to solve the frw flatness problem , let s see what normal inflation would do to solve the bulk flatness problem .
let s extend the toy model of the previous section by allowing a brane scale factor @xmath103 : @xmath104 where again , @xmath40 is the extra dimension coordinate and our flat brane is located at @xmath105 .
it is easy to see that no matter how much we inflate our brane by arranging @xmath103 to grow exponentially , we will not recover the @xmath0 isometry
. one possible resolution to this problem is to inflate the @xmath40 dimension .
the difficulty lies in inflating the extra dimensions to smooth out the @xmath0 invariance violating curvature while preserving the large warping . to see this ,
introduce a bulk scale factor @xmath106 for the extra dimension .
the toy metric is then @xmath107 inflating @xmath106 ( `` bulk inflation '' ) will render the curvature set by @xmath95 harmless .
one can see this by making the coordinate transformation @xmath108 , in which case the metric becomes @xmath109dt^{2 } - a^{2}(t ) \
e^{-2b\tilde{u}/c}\ d\vec{x}^{2 } -d\tilde{u}^{2 } + 2 \left(\frac{\dot{c}}{c}\right)\tilde{u}\ d\tilde{u } \ , dt .
\label{eq : inflated}\ ] ] now imagine @xmath110 inflates by large amount , after which @xmath111 ( or at least @xmath112 )
. then the factor @xmath113 , which sets the spatial curvature scale in the bulk , can be made arbitrarily small while still recovering @xmath0 isometry .
of course the price one pays here is that inflation of the bulk has inflated away the warp factor !
an obvious loophole in the argument thus far is the possibility of a hierarchy between the lorentz violating curvature and the warp factor curvature .
then one may be able to inflate away the lorentz violating curvature without erasing the warp factor .
for example , if one complicates the toy metric a bit further and takes it to be @xmath114 - c^{2}(t )
\ du^{2}\ ] ] with @xmath115 , then bulk inflation may dilute away the curvature due to @xmath95 while maintaining the warp factor from @xmath116 .
the challenge then is to come up with a physical scenario with a large hierarchy between @xmath116 and @xmath95 .
of course , there may be other solutions that involve the bulk inflating and then shrinking in such a way that the @xmath0 violating curvature is removed , but the main point still stands : inflationary model building has a new problem to solve . finally , consider the large compact extra dimension scenarios @xcite with @xmath67 extra spatial dimensions .
imagine that an @xmath0 violating metric of the form eq .
( [ eq : beforeinflation ] ) has been inflated as in eq .
( [ eq : inflated ] ) to end with an acceptably flat , compact extra dimension .
let us parameterize such a metric as @xmath117 where @xmath28 correspond to the compact extra dimensions where @xmath118 runs through 4 to @xmath119 .
if the extra dimensions described by the coordinates @xmath29 are compact , then inflation through @xmath106 poses the danger of the compact dimensions being initially too small to be described by an effective field theory .
more specifically , let us define the total expansion of the extra dimensions between some initial time @xmath120 and today ( @xmath35 ) as @xmath121 where the second equality is required by the requirement that the curvature not violate @xmath0 invariance out to scale of @xmath122 .
suppose the size of the extra dimensions is initially @xmath123 .
after inflation , the size of the extra dimensions is @xmath124 .
then since the four - dimensional planck scale requires @xmath125 where @xmath118 is the @xmath67-dimensional planck scale , we find @xmath126 if we require an effective field theory description to be valid by imposing @xmath127 , we find @xmath128 this means that if we require that the fundamental planck scale @xmath118 to be @xmath129 tev , we find that the curvature in the bulk before bulk inflation can be only very tiny , @xmath130ev .
hence , we find that for large _ compact _ extra dimensions , only very tiny curvatures can be smoothed out by inflation .
this suggests that a non - quantum - gravitational theory of inflation can not smooth away the bulk curvature , and most likely some fundamental symmetry must play a role in initially setting the @xmath0 violating curvature to zero .
a related discussion can be found in ref .
@xcite .
let us reiterate the main point of this section which is the main result of this paper . for _ noncompact
_ warped extra dimension scenarios , one must invoke a mechanism such as inflation to smooth out the @xmath0 isometry violating curvature to obtain a lorentz invariant effective field theory .
however , the difficulty with this resolution is that in such an inflationary scenario , the bulk warp factor that one needs to localize gravity may be inflated away altogether .
hence , the new flatness problem is to inflate away the @xmath131 violating curvature selectively while preserving a large warp factor . in the case of large compact extra dimensions ,
we find that inflation at the level of an effective field theory generically can not make the bulk flat .
we have shown that curvature in the bulk leads to an apparent breaking of lorentz symmetry with respect to an observer living on the brane observing a field propagating in higher dimensions . unlike the @xmath0 violation in the four - dimensional world due to curvature , lorentz violation in theories with fields propagating in higher dimensions persists as long as the spacetime does not possess an @xmath0 isometry , no matter how large of an energy the four - dimensional effective field theory is probed .
this is in contrast with what is dictated by the equivalence principle in four dimensions .
we called this apparent uv - limit - persisting violation of four - dimensional equivalence principle for dimensionally reduced theories the `` inequivalence principle . ''
note that lorentz violation for the dimensionally reduced theory is true even when the brane has an @xmath0 isometry .
furthermore , the mismatch between the brane isometry group and the full spacetime isometry group results in an ambiguous geometrical picture from a four - dimensional empirical point of view regardless of the symmetry group .
this implies a new flatness problem for cosmological scenarios having large ( possibly noncompact ) extra dimensions . for warped noncompact
extra dimensions scenarios , the problem is to come up with a mechanism to flatten the @xmath0 violating wrinkles while preserving the large warp factor necessary for graviton trapping .
inflation generically helps to smooth away the wrinkle , but it also eliminates the warp factor . for flat compact large extra dimensions scenarios , we demonstrated that inflation at the effective field theory level is not sufficient to smooth away any significant curvature in the bulk .
it is tempting to speculate that the first signatures for large extra dimensions may come from deducing the existence of lorentz violations in the early universe .
this would be possible only if the observable anomalies in the early universe arising from lorentz violations are sufficiently distinct from the anomalies arising from other effects .
indeed , since in most popular scenarios only gravity and other extremely weakly interacting ( with the sm fields ) fields propagate in the bulk , it may be difficult to observe the lorentz violations with respect to the zero modes of these fields unless the violation is extreme .
one lorentz violating observable characteristic of the inequivalence principle is the wavelength independent deviation of signal propagation speed .
for example , for the free scalar field of eq .
( [ 5dscalar ] ) propagating in a five - dimensional background of eq .
( [ metric ] ) , one can solve the wave equation perturbatively , perturbing with the parameter @xmath132 where @xmath95 characterized the lorentz violating curvature and @xmath21 is the characteristic size of the extra dimension .
one can write one of the modes as @xmath133 where @xmath134,\ ] ] and @xmath135
this implies that the group velocity is @xmath136 in agreement with eq .
( [ dimreduced ] ) since there we have @xmath137 as we have argued before , the propagation speed of gravitational waves will be similar .
hence , we conclude that one may be able to constrain the bulk - curvature violating @xmath0 isometry by comparing the gravitational wave arrival time and the light arrival time .
imagine measuring the time correlation of the arrival of the gravitational wave and light pulses from a gamma - ray burst .
taking the gamma - ray burst to be at a distance @xmath138 mpc @xmath139 s , and assuming a resolution for the arrival times of the pulses of @xmath140 s , we would be able to constrain the lorentz violating curvature to be smaller than @xmath141 if no time lag is detected . in conclusion , any realistic cosmological model with extra dimensions must account for a mechanism to generate approximate lorentz symmetry for fields in the bulk , if the dimensionally reduced bulk field such as the graviton can be shown to be approximately lorentz invariant .
if we are lucky , perhaps nature will give us a clue regarding the extra dimensions through tests of lorentz violations in the graviton sector and any other sector that may live in extra dimensions .
djhc thanks l. everett , g. l. kane , j. wang , j. liu , j. lu , and r. garisto for interesting conversations .
djhc was supported by the department of energy .
ewk was supported by the department of energy and nasa under grant nag5 - 7092 .
if the universe had even a tiny positive large - scale spatial curvature at some early time , it would have collapsed without expanding sufficiently . without sufficient expansion , the energy density would never have sufficiently diluted to allow the curvature of spacetime to be small compared to the electroweak scale
. for discussions about the role of d - branes at early epochs , see e.g. m. maggiore and a. riotto , nucl .
b 548 , 427 ( 1999 ) [ hep - th/9811089 ] ; a. riotto , phys . rev .
d 61 , 123506 ( 2000 ) [ hep - ph/9904485 ] .
this exact form of this _ static _
metric is most likely unphysical because one can easily show that it can be derived from a relativistic fluid with an equation of state @xmath143 embedded in an ads space .
however , for illustrative purposes this is a useful , simply calculable toy model
. we will later give a slightly more realistic example .
although the details resulting from the spin - two nature and the coupling to sources will change the details between the graviton propagator and the scalar field propagator , the qualitative behavior of the field propagation should be the same . | there is no known fundamental reason to demand as a cosmological initial condition that the bulk possess an @xmath0 isometry . on the contrary ,
one expects bulk curvature terms that violate the @xmath0 isometry at early epochs , leading to a violation of lorentz invariance on our brane . demanding that the lorentz noninvariant terms are small leads to a new `` flatness '' problem , not solved by the usual formulation of inflation .
furthermore , unlike in four dimensions , the lorentz violations induced from the bulk curvature can not always be removed as the infrared cutoff is taken arbitrarily large .
this means that the equivalence principle in higher dimensions does not guarantee the equivalence principle in dimensionally reduced theories .
near - future experiments are expected to severely constrain these lorentz - violating `` signatures '' of extra dimensions . 0.0 in |
the top quark is the heaviest known elementary particle of the standard model ( sm ) .
it has a life - time shorter than the hadronization time and offers a window to search for new physics . with large numbers of top events produced at lhc in run-1
, it is possible to study its properties in detail .
these proceedings present measurements of top production , top properties and searches at the center of mass energy of 7 , 8 and 13 tev by the atlas experiment @xcite .
the atlas experiment has performed measurements of the inclusive top - antitop production cross section of @xmath3 production at center of mass energy of 7 , 8 and 13 tev in the dilepton channel @xcite .
the measurement at 8 tev represents the most precise measurement to date , even more precise than next - to - next - to - leading order ( nnlo ) predictions @xcite . at 13 tev
, the cross section is found to be @xmath4 . in the three analyses , a simultaneous fit of the cross section
, b - jet reconstruction and tagging efficiency is performed in the @xmath5 channel .
all the measurements are in agreement with sm predictions , as shown in figure [ fig : a ] . 0.50 pair production in pp collisions as a function of centre - of - mass energy .
the results at @xmath6 , @xmath7 and @xmath8 tev are compared to the theory predictions @xcite .
( b ) summary of atlas measurements of the single top production cross sections in various channels as a function of the center of mass energy compared to theoretical calculations @xcite.,title="fig : " ] 0.40 pair production in pp collisions as a function of centre - of - mass energy .
the results at @xmath6 , @xmath7 and @xmath8 tev are compared to the theory predictions @xcite .
( b ) summary of atlas measurements of the single top production cross sections in various channels as a function of the center of mass energy compared to theoretical calculations @xcite.,title="fig : " ] new physics could affect the shape of the differential cross section as a function of top kinematics . in this conference , measurements performed in two different topologies were presented : the resolved and the boosted topologies , which are optimized for tops with transverse momentum @xmath9 and and @xmath10 gev respectively . in the resolved topology , the measurement of the top - antitop differential cross section is performed as a function of the mass , @xmath11 and rapidity of the top pair system at @xmath8 tev @xcite .
these measurements are performed in a fiducial region closely matching the detector acceptance and using tops observables based on stable particles in the @xmath5 channel .
r0.4 the observed spectra are softer than the monte carlo ( mc ) predictions . in the boosted topology ,
the differential cross section as a function of the @xmath11 of the @xmath3 system is measured in the semileptonic decay channel at @xmath7 tev @xcite .
the measurements are performed in both a fiducial and total phase space .
the measurements are also found to be softer than the mc predictions .
several measurements of the inclusive single top cross section in the @xmath12 , @xmath13 and @xmath14 channel at 7 and 8 tev have been performed by atlas .
the latest 8 tev results were presented in this conference .
figure [ fig : b ] shows a summary of these measurements @xcite .
the inclusive cross sections for @xmath13-channel is found to be @xmath15 pb , showing a good agreement with the nlo+nnll predictions @xcite , as well as the inclusive measurement in the @xmath14 channel where the measured value is @xmath16 pb @xcite .
an upper limit for the @xmath12-channel cross section was established @xcite .
the @xmath17 and @xmath18 cross sections were measured in final states with two , three or four leptons @xcite . in this analysis , a simultaneous fit is performed in 20 signal and control regions .
the control regions are defined by varying the number jets and lepton charge in the event .
the results agree with the sm prediction as shown in figure [ fig : ttwttv ] .
measurements of the top mass , performed in @xmath3 events in the semileptonic and the dileptonic channels , were presented @xcite . the analysis in the semileptonic channel uses a 3-dimensional template fit which determines simultaneously the top quark mass , the jet and @xmath19-jet energy scale factor .
this measurement is the most precise mass measurement performed by atlas to date .
the dilepton channel uses a one dimensional template fit and the combined result is found to be @xmath20 gev . in single
top decays , the measurement is performed through a one dimensional template fit using the invariant mass of the lepton and @xmath19-jet , in order to reduce the modelling uncertainties @xcite .
neural networks are used to optimize the purity of the selection to @xmath21 .
the measured value is found to be @xmath22 .
moreover , a measurement of the pole mass using the normalized differential cross section for @xmath23-jet as a function of the inverse of the invariant mass of system was performed @xcite .
the selected events are identified using the lepton+jets top - quark pair decay channel , where lepton refers to either an electron or a muon .
the measured value is found to be @xmath24 .
top pairs produced via the strong interaction are produced unpolarized at leading order .
their spins are correlated and information transferred to decay products .
a measurement of the @xmath3 spin correlation at @xmath7 tev was performed @xcite . in this analysis ,
the @xmath3 spin correlation is extracted from dilepton events by using the difference in the azimuthal angle of the two selected leptons .
the measured value is consistent with the sm predictions .
moreover , the measurement of the angular distribution of the decay products is sensitive to super symmetric top squark ( stop ) pair production .
the top squarks with masses between the top quark mass 172.5 gev and 191 gev are excluded at the @xmath25 confidence level .
in this conference , limits in searches for flavor changing neutral currents ( fcnc ) were presented @xcite .
fcnc are predicted by the sm but are largely suppressed by the gim mechanism . in beyond the standard model ( bsm ) models the fcnc are enhanced , giving a window to search for new physics .
one search is performed in @xmath3 events with one top quark decaying to a @xmath26-boson and a light quark and the other to a @xmath27-boson and a @xmath19-quark .
both bosons are required to decay leptonically .
no evidence of fcnc is found and an observed upper limit is established .
another search for fcnc is performed in single top production via @xmath28 with the top decaying to a @xmath26-boson and a light quark @xcite .
only the leptonic channel is considered and upper limits are established on the cross section times branching ratio and on the coupling constants of the processes .
atlas has performed multiple measurements in the top physics domain .
the measurements of cross sections and properties are in agreement with the sm predictions .
some top quark properties were used to exclude bsm models and limits were established for fcnc searches .
99 atlas collaboration , 2008 jinst 3 s08003 .
atlas collaboration , eur .
j. c 74 ( 2014 ) , no .
10 , 3109 .
m. czakon , p. fiedler and a. mitov , phys .
lett . 110 ( 2013 ) , 252004 . atlas collaboration , atlas - conf-2015 - 033
. atlas collaboration , jhep 1506 ( 2015 ) , 100 . | these proceedings summarize the latest measurements on top production , top properties and searches using the atlas detector at the lhc .
the measurements are performed on @xmath0 collision data with a center of mass energy @xmath1 and @xmath2 tev . |
electric vehicles ( evs ) have several advantages over the traditional gasoline powered vehicles . for example , evs are more environment friendly and more energy efficient . realizing the above
, regulators ( e.g. federal energy regulator commission ( ferc ) ) are providing incentives to the consumers to switch to electric vehicles .
manufacturers ( e.g. tesla , nissan ) are increasingly developing evs equipped with superior technologies . as a result , electric vehicles
are increasingly become popular .
however , a wide deployment of evs requires an extensive network of charging stations which can be capable of charging large number of vehicles .
vehicle - to - grid ( v2 g ) service has been proposed @xcite to enhance the profitability of the evs . in the v2 g service
, evs can inject energies to the grid by discharging from their batteries .
thus , this bidirectional power flow where evs can both charge and discharge has a lot of potential .
hence , a lot of effort is going on for developing bidirectional evs@xcite .
however , the higher amount of charging and discharging cycles will degrade the battery life . hence , the owners of the evs have to be compensated adequately for the v2 g service .
thus , though a charging station can gain an additional profit using the v2 g service of the evs , however without a proper pricing mechanism , the owners of the evs will not prefer the v2 g service in the first place which may nullify the profit of the charging station . without a proper control mechanism , the cost of the charging station and the peak energy consumption may increase . without a profitable charging station , the wide - scale deployment of the evs will remain a distant dream .
the charging station needs to select prices in order to earn profits by allocating resources in an intelligent manner among the evs .
the charging station also needs to provide adequate compensation to the owners if the ev is used for the v2 g service .
however , high prices or low compensation may not provide incentives to the owners which may reduce the profit .
hence , a proper pricing mechanism for charging the evs and the v2 g service is imperative for a charging station .
we propose a menu based pricing scheme for charging an ev .
whenever an ev arrives at the charging station , the charging station offers a variety of contracts @xmath0 to the ev s owner ( or , user ) at a price @xmath1 to the user where the user will be able to charge at least @xmath2 units of energy within the deadline @xmath3 for completion , and the battery usage will be limited to @xmath4 amount .
the battery usage is the total amount of charging and discharging of the ev .
the user either accepts one of the contracts by paying the specified price or rejects all of those based on its payoff .
we assume that the user gets a utility for consuming @xmath2 amount of energy within the deadline @xmath3 .
however , the user also has to incur a cost for battery utilization @xmath5 .
the payoff of the user ( or , user s surplus ) for a contract is the difference between the utility and the sum of the cost incurred , and the price to be paid for the contract .
the user will select the option which fetches the highest payoff .
the various advantages of the above pricing scheme should be noted .
first , it is an online pricing scheme .
it can be adapted for each arriving user .
second , since the charging station offers prices for different levels of charging required and the deadline , the charging station can prioritize one contract over the others depending on the energy resources available .
third , the charging station also provides options of the maximum battery utilization to the users .
the user who is not interested in the v2 g service is entitled to do that by selecting the contract with @xmath5 as @xmath6 .
finally , the user s decision is much simplified .
she only needs to select one of the contracts ( or , reject all ) .
we consider that the charging station is equipped with renewable energy harvesting devices and a storage device for storing energies .
the charging station may also buy conventional energy from the market to fulfill the contract of the user if required .
hence , if a new user accepts the contract @xmath0 , a cost is incurred to the charging station .
this cost may also depend on the existing evs and their resource requirements .
a contract also specifies the maximum battery utilization which restricts the v2 g service generated from an ev .
hence , the charging station needs to find the optimal cost for each contract .
we show that obtaining the cost of fulfilling a contract is a _ linear programming _ problem .
we also show that if a user accepts a contract with a higher battery utilization , the cost of the contract is lower ( lemma 1 ) .
we consider two optimization problems
i ) social welfare maximization , and ii ) the ev charging station s profit maximization .
we investigate the existence of a pricing mechanism which maximizes the ex - post social welfare , _
i.e. _ , maximizes the social welfare for every possible realization of the utility function .
we show that there exists such a pricing strategy .
the pricing scheme is simple to compute , as the charging station selects a price which is equal to the marginal cost for fulfilling a certain contract for a new user ( theorem [ thm : max_socv2 g ] ) .
however , the above pricing scheme only provides _
zero _ profit to the charging stations .
thus , such a pricing scheme may not be useful to the charging station .
we show that when a charging station is _ clairvoyant _
( i.e. , the primary knows the utilities of the users ) , there exists a pricing scheme which satisfies both the objectives ( lemma [ thm : profit_max ] ) .
though in the above pricing mechanism , the user s surplus becomes @xmath6 .
a _ clairvoyant _
charging station may not be beneficial for the user s surplus . in the scenario where the charging station does not know the exact utilities of the users ,
we show that _ there may not exist a pricing strategy which simultaneously maximizes the ex - post social welfare and the expected profit_. one has to give away the ex - post social welfare maximization in order to achieve expected profit maximization .
_ however , the user s surplus becomes higher compared to the clairvoyant scenario_. hence , an uncertainty of the utility enhances the user s surplus .
we propose a pricing strategy which can fetch the highest possible profit to the charging station under the condition that it maximizes the ex - post social welfare ( theorem [ thm : profitmax_uncertainty ] ) . above pricing strategy
provides a _ worst case _ maximum profit to the charging station . since the above pricing strategy may not yield the
_ maximum expected _ profit to the charging station , we have to relax the constraint the social welfare to be maximized in order to yield a higher profit to the charging station .
whether a contract will be selected by the user does not depend on the price of the contract , but also , the prices of other contracts .
thus , achieving a pricing scheme which maximizes the expected profit is difficult because of the discontinuous nature of the profits .
we propose a pricing strategy which yields a fixed ( say @xmath7 ) amount of profit to the charging station .
further , we show that a suitable choice of @xmath7 can maximize the profit of the charging station for a class of utility functions ( theorem [ thm : aclassutility ] ) . in section [ sec : v2g_profitability ]
we characterize the conditions which will yield higher profits to the charging station in the v2 g service .
we show that if the conventional energy price is high , the charging station selects more preferable prices for enticing v2 g services if the renewable energy harvested is low . when the charging station s storage capacity is low and the renewable energy harvesting is low , the user s incentives for the v2 g service also increases .
the v2 g service also increases the profit of the charging station .
finally , we , empirically provide insights how a trade - off between the profit of the charging station and the social welfare can be achieved for various pricing schemes ( section [ sec : simulation_results ] ) .
our numerical analysis shows that both the user s surplus and the charging station s profit increase with the v2 g service . the energy consumption during the peak period decreases as users provide more v2 g services during the peak period .
charging schedule for evs using price signals for _ unidirectional _ service have been considered@xcite .
the above papers did not consider the optimal discharging schedule of the evs as these papers only considered unidirectional power flow . as a result
, these paper did not consider the battery degradation cost incurred in the v2 g service .
few papers considered the _ bidirectional _ power flow @xcite . however , their focus was scheduling of the charging and discharging pattern of the evs , rather than the pricing aspects .
naturally , these papers did not consider whether users will prefer the amount of battery degradation found in the optimal scheduling process .
the social welfare maximizing and profit maximizing prices are also not considered .
@xcite considered the optimal pricing to the evs in a day - ahead setting for residential charging .
the users control the charging and discharging pattern at each instance for the price selected by the aggregator in a pre - specified manner .
however , the users can arrive randomly in the charging station , the charging station needs to select a price for each arriving user using an online algorithm .
the users can not control the charging and discharging schedule at each instance in the charging station . compared to _ all _ the above papers , in our approach , the charging station specifies the amount of battery utilization for each contract .
the charging station also selects prices for different deadlines in our approach . hence , the user can now choose its own deadline and can specify the v2 g service it is willing to commit based on its own need .
deadline differentiated prices have been considered @xcite . in our previous work @xcite
, we also considered a menu - based pricing scheme where the charging station provides prices for different deadlines and prices to the new user .
however , the above papers did not consider the _ bidirectional _ power flow problem .
thus , in this setting the charging station now also needs to select different prices for different amounts of battery utilization .
_ to the best of our knowledge , this is the first attempt to consider menu - based pricing which considered the bidirectional power flow from the evs . _
further , this is the first work that incentivizes the users to participate in the v2 g service by finding optimal prices for different amounts of battery degradation while maximizing the social welfare or profit of the charging station .
we consider that evs arrive throughout a day at the charging station for charging .
suppose that the user @xmath8 arrives at time @xmath9 . the job of the charging station is to select a price for charging in order to maximize the profit .
we consider a vehicle - to - grid ( v2 g ) service where electric vehicles ( evs ) can feed back stored power to the grid .
specifically , the energy can be discharged from the batteries of the evs . however , the ev batteries have fixed number of charging - discharging cycles @xcite . when the battery is used for feeding back energy to the grid
, the battery wear cost may be significant .
thus , the users may want to limit the battery utilization as low as possible .
we consider that a charging station wants to maximize its profit over a certain time period ( e.g. over a day ) .
it will offer a menu - based price contract @xmath10 for contract @xmath0 to the user @xmath8 which arrives at time @xmath9 in the charging station ( fig .
[ fig : ev_charging ] ) .
if the user selects the menu @xmath0 , then , the user @xmath8 will be able to charge @xmath2 amount of energy at most within the deadline @xmath3 and the maximum amount of _ additional _ battery utilization ( definition 1 ) is restricted to @xmath5 .
the deadline @xmath3 denotes that the user has to leave by time @xmath3 , after which the ev will not be able to charge .
first , we describe the constraints the charging station has to satisfy the contract @xmath11 for the user @xmath8 .
* charging / discharging rate * : let @xmath12 be the amount of energy provided to ( or , discharged from ) the ev @xmath8 during time @xmath13 .
@xmath14 indicates that the ev @xmath8 is charged and @xmath15 indicates that energy is discharged from the ev @xmath8 during time @xmath13 .
also note that there is an initial set @xmath16 of existing evs .
vehicle @xmath17 requires additional @xmath18 amount of energy within the deadline @xmath19 .
the charging and discharging efficiency of the ev @xmath8 is denoted as @xmath20 and @xmath21 respectively .
note that @xmath22 where @xmath23 is the positive part of @xmath12 ( it denotes that the electric vehicle @xmath8 is charged during time @xmath13 ) and @xmath24 denotes the negative part of @xmath12 ( it denotes the amount discharged from ev @xmath8 during time @xmath13 ) .
hence , the following set of constraints must be satisfied @xmath25 @xmath26 indicates the total amount of energy required for charging and @xmath27 indicates the total amount of energy used when evs discharge during time @xmath13 .
note that an ev can not simultaneously charge and discharge , hence , we must have @xmath28 for all @xmath29 and @xmath3 . _
later we will show in theorem [ thm : noposinegative ] that in an optimal solution @xmath28 though we have not explicitly considered the above constraint in the system .
* ev s battery limit * : let the battery level of the ev @xmath30 at time @xmath3 be @xmath31 . let the battery level at the start of the time @xmath9 be @xmath32 . since the battery of the ev can be charged or discharged in the v2 g service , the total amount of charging and discharging must satisfy the limits on the battery levels : the battery level must be between a lower value @xmath33 ( it can be @xmath6 ) and the high value @xmath34 ( the highest capacity ) . hence , we must have @xmath35 note that @xmath36 is @xmath37 . * charging and discharging rate limit * : there is also a charging and discharging limit . hence , we have @xmath38 @xmath39 , and @xmath40 are respectively the charging and discharging rate limits . *
hybrid source * : the charging station is equipped with renewable energy harvesting devices ( fig .
[ fig : hybrid_v2 g ] ) .
the charging station also has a storage device with capacity @xmath41 .
the charging station has the forecast of harvested energy as @xmath42 for time @xmath13 .
the charging station also can buy energies from the conventional market .
suppose that the amount of energy used from the storage device during time @xmath13 be @xmath43 and let the energy be stored from the electric vehicles and the conventional energy in the storage device be @xmath44 during time @xmath13 ( fig .
[ fig : hybrid_v2 g ] ) .
let @xmath45 be the level of battery at time @xmath3 .
the charging efficiency and discharge efficiency of the battery of the charging station is considered to be @xmath46 and @xmath47 respectively .
thus , @xmath48 note that our model can also incorporate the scenario where the batteries have some static leakage rate i.e. the battery level decreases with time .
* energy to and from the grid * : let the amount the charging station buys from the conventional market and sells to the grid as @xmath49 and @xmath50 respectively for time @xmath13 ( fig : [ fig : hybrid_v2 g ] ) .
hence , we have @xmath51 since @xmath44 is the negative of @xmath43 , thus , we can represent the above constraint in the following @xmath52 we also must have @xmath53 since the battery of the charging station should not charge or discharge at the same time . though we have not explicitly considered the above constraint , however , we show that in an optimal solution in theorem [ thm : noposinegative ] , we always have @xmath53 .
note that the constraints in ( [ eq : battery_capacity ] ) and ( [ eq : storage ] ) also specify the bound on the amount of charging and discharging from the evs .
the amount of charging can not exceed the total amount of stored energy in the battery and the energy bought from the grid .
the stored energy depends on the harvested renewable energy , energy charged to the evs and discharged from the evs , energy bought and energy sold to the grid .
[ fig : hybrid_v2 g ] depicts various system parameters .
* maximum battery utilization * in the vehicle - to - grid ( v2 g ) service , frequent charging and discharging may reduce the lifecycle of the battery of an ev .
thus , the users may not like the evs be charged and discharged very often .
we define a metric which will model the total maximum utilization of the battery of an ev .
battery utilization is defined as the absolute value of the difference between the battery levels at two subsequent time intervals .
hence , if the deadline is @xmath54 for user @xmath8 , then the total battery utilization for user @xmath8 is @xmath55 .
the battery utilization defines the total level of charging and discharging has been done . if the ev @xmath8 is used only for charging , then @xmath56 . in a contract @xmath57
, the battery of ev needs a charging amount of @xmath2 , and thus , the total battery utilization has to be at least @xmath2 .
thus , the contract @xmath11 where @xmath58 denotes the _ additional battery utilization _ apart from the charging amount @xmath2 specified by the contract .
we denote @xmath5 as the maximum _ additional _ battery utilization with slight abuse of notation .
specifically , in the contract @xmath11 , the maximum utilization is restricted to @xmath4 for user @xmath8 .
note that @xmath2 is the energy that the user @xmath8 will receive in the contract @xmath11 .
suppose the maximum utilization remaining for an existing user @xmath17 at time @xmath9 is @xmath59 .
thus , if the user @xmath8 selects the contract @xmath11 , the constraint that the charging station has to satisfy is @xmath60 the above constraint is not linear . in the following
we reduce it in a linear form .
note from ( [ eq : dischargelimit ] ) that @xmath61since @xmath62 and @xmath63 are the positive and negative parts of @xmath37 , thus , @xmath64 .
hence , the expression in ( [ eq : bu ] ) becomes @xmath65 note that we must have @xmath28 .
we show in theorem [ thm : noposinegative ] that this is indeed true in an optimal solution .
user s utility for the contract @xmath0 is denoted as @xmath66 which is a random variable .
the realized value @xmath67 is only known to the user @xmath8 , but not known to the charging station in general .
the payoff of user @xmath8 or _
user s surplus _ if she selects the contract @xmath0 is @xmath68 ( fig .
[ fig : ev_charging ] ) . if she rejects all her payoff is @xmath6 .
the user will select the contract that fetches the maximum payoff to her .
thus , for a menu of prices @xmath10 , the user @xmath8 selects @xmath69 $ ] such that it maximizes the following @xmath70 note that @xmath71 only if @xmath72 .
if such a solution is not unique , any convex combination of these solutions is also optimal since a user can select any of the maximum payoff contracts .
we denote the decision as @xmath73 .
note that the decision whether to accept the menu price @xmath10 depends not only on the price @xmath10 but also other price menus i.e. @xmath74 where @xmath75 and @xmath76 and @xmath77 .
this is because the user only selects the price menu which is the most favorable .
note that if the maximum payoff that user gets among all the price menus ( or , contracts ) is negative , then the user will not charge i.e. , @xmath78 for all @xmath2 , @xmath3 and @xmath79 .
we also assume that if there is a tie between charging and not charging , then the user will decide to charge i.e. , if the maximum payoff that user can get is @xmath6 , then the user will decide to charge . amount . ] since the users arrive to the charging station at any time throughout the day , the charging station does not know the exact arrival times for the future vehicles .
we consider that the charging station is myopic or near - sighted i.e. , it selects its price for user @xmath8 without considering the future arrival process of the vehicles .
however , it will consider the cost incurred to charge the existing evs .
note that as the number of existing users increases , the marginal cost can increase to fulfill a contract for an arriving user , hence , such a pricing strategy may not maximize the payoff in a long run .
note that , _ a myopic pricing strategy is optimal in the case the marginal cost of fulfilling a demand of a new user is independent of the number of existing users .
_ in practice , the charging station often has fixed number of charging spots .
thus , the charging station may want to select high prices for user @xmath8 , in order to make the charging spots available for the users who can pay more but only will arrive in future . hence , considering the future arrival process of the vehicles may increase the profit of the charging station
. however , such a pricing strategy is against the first come first serve basis which is the current norm for charging vehicles .
our approach can be seen as a fair allocation process , where the charging station serves users based on the first come first serve basis .
the optimal cost for the charging station to fulfill the contract @xmath57 to the user @xmath8 is thus @xmath80 where @xmath81 is the per unit cost of buying energy from the grid and @xmath82 is the per unit cost of selling energy to the grid . note that our model can also incorporate time varying , strictly increasing convex costs @xmath83 or concave prices @xmath84.we assume that @xmath85 in order to avoid arbitrage opportunity .
however , the above cost structures will destroy the properties of the linear programming .
our model can be easily extended to the setting where there are upper bounds on @xmath49 and @xmath50 .
the problem will still be a linear programming problem .
the cost @xmath81 and the price @xmath82 are assumed to be known . in case they change in a dynamic manner i.e. , they are real time prices , we consider that @xmath81 is the expected cost , and @xmath82 is the expected price .
note that the above problem is a linear programing problem ( similar to the setting considered in @xcite ) and thus , it is easy to solve .
however , the number of constraints are hugher compared to @xcite as we consider the v2 g service . in the following theorem
, we show that the optimal cost to the charging station given in ( 12 ) , @xmath86 and @xmath53 for all @xmath3 , and @xmath87 .
[ thm : noposinegative ] in an optimal solution of the problem @xmath88 , we must have the decision variables @xmath28 @xmath89 and @xmath3 .
further , in an optimal solution , we also have @xmath53 for all @xmath3 .
_ proof _ : see appendix [ sec : proof_thmnoposinegative ] .
the above theorem shows that in an optimal solution , neither the ev nor the battery of the charging station simultaneously charge and discharge .
now we calculate the _ additional _ cost imposed to the charging station for fulfilling the new contract .
first , we introduce some notations which we use throughout .
[ defn : vlt ] the charging station has to incur the cost @xmath90 for serving the existing users and the contract @xmath11 for the new user @xmath8 where @xmath90 is the value of the linear optimization problem @xmath88 .
since @xmath88 is a linear optimization problem , it is easy to compute @xmath91 .
further , note that if the above problem is infeasible for some @xmath2 , @xmath3 and @xmath5 , then we consider @xmath91 as @xmath92 .
[ defn : v - k ] let @xmath93 be the amount that the charging station has to incur to satisfy the requirements of the existing evs if the new user does not opt for any of the price menus .
if user @xmath8 does not accept any price menu , then the charging station still needs to satisfy the demand of existing users i.e. , the charging station must solve the problem @xmath88 with @xmath94 .
@xmath93 is the value of that optimization problem . from definitions [ defn : vlt ] and [ defn : v - k ]
we can visualize @xmath95 as the additional cost or marginal cost to the charging station when the user @xmath8 accepts the price menu @xmath96 .
we assume that the prediction @xmath97 is perfect for all future times and is known to the charging station .
however , menu - based pricing approach can be extended to the setting where the estimated generation does not match the exact amount .
first , we can consider a conservative approach where @xmath42 can be treated as the worst possible renewable energy generation . as a second approach
, we can accumulate various possible scenarios of the renewable energy generations , and try to find the cost to fulfill a contract for each such scenario .
for example , if there are @xmath98 number of possible instances of the renewable energy generation amount in future .
then , we can find the optimal cost for each such instance of renewable energy generation @xmath99 where @xmath100 instead of @xmath42 .
we then can compute the average ( or , the weighted average , if some instance has greater probability ) of the optimal costs , and that cost can be taken as the cost of fulfilling a certain contract .
the following result entails that the v2 g service indeed reduces the cost of the charging station .
[ lm : v2 g ] @xmath101 for any @xmath102 .
thus , @xmath103 for any @xmath104 .
_ outline of the proof _ : note from ( [ eq : battery_utilization ] ) that the decision space is more restricted for @xmath105 compared to @xmath106 if @xmath102 . every feasible solution with @xmath105 is also feasible with @xmath106 since @xmath102 .
thus , the above result follows . since the charging station can sell some energies taken from the evs , the cost of fulfilling the contract @xmath0 will be lower as @xmath5 will increase .
if the user @xmath8 selects the menu @xmath0 , then the additional cost incurred by the charging station is @xmath107 . hence , the profit of the charging station is @xmath108 .
if the user does not select any of the menus , then the charging station gets @xmath6 profit .
thus , the expected profit of the charging station is @xmath109 where @xmath110 is the probability of the event that the menu @xmath0 is accepted by the user @xmath8 . note from ( [ eq : profitmax_expec ] ) that the profit maximization is a difficult problem as the menu selected by an user inherently depends on the prices selected for other menus .
for example , if the price selected for a particular contract is high , the user will be reluctant to take that as compared to a lower price one .
the profit is a discontinuous function of the prices and thus , the problem may not be convex even when the marginal distribution of the utilities are concave .
we consider that the charging station decides the price menus in order to fulfill one of the two objectives ( or , both) i ) social welfare maximization and ii ) its profit maximization .
the social welfare is the sum of user surplus and the profit of the charging station .
as discussed in section [ sec : userutility ] for a certain realized values @xmath67 if the user @xmath8 selects the price menu @xmath10 , then its surplus is @xmath68 , otherwise it is @xmath6 . as discussed in section
[ sec : profit ] the profit of the charging station is @xmath111 for a given price @xmath10 if the user selects the menu , and it is @xmath6 otherwise .
hence , the social welfare maximization problem is to select the price menu @xmath112 which will maximize the following @xmath113 recall that in order to find @xmath91 we have to solve @xmath114(cf .
( [ eq : vlt ] ) ) which is a constrained optimization problem .
since the charging station is unaware of the utilities of the users , the charging station has two choices - i ) decides a price and hopes that it will maximize the social welfare for the realized values of utilities ( _ ex - post _ maximization ) , or ii ) decides a price and hopes that it will maximize the social welfare in an expected sense ( _ ex - ante _ maximization ) .
thus , the _ ex - ante _ maximization does not guarantee that the social welfare will be maximized for every realization of the random variables @xmath115 .
however , in the _ ex - post _ maximization , the social welfare is maximized for each possible realization of the random variables .
thus , _ ex - post _ maximization is a stronger concept of maximization ( and thus , also more desirable ) and it is not necessary that there exist pricing strategies which maximize the ex - post social welfare .
however , we show that in our setting there exist pricing strategies which maximize the ex - post social welfare .
note that _ ex - post _ social welfare maximization is the same as ( [ eq : socialwelfare_maxproblem ] ) .
social welfare maximization does not guarantee that the charging station may get a positive profit .
it is important for the wide scale deployment of the charging stations that the charging station must have some profit .
the charging station needs to select @xmath10 in order to maximize the expected profit ( given in ( [ eq : profitmax_expec ] ) ) .
note that in order to select optimal @xmath10 , the charging station has to obtain @xmath91 i.e. , it has to solve the problem @xmath116 ( cf .
( [ eq : vlt ] ) ) for each choice of @xmath2 , @xmath3 and @xmath5
. note that in order to select optimal @xmath10 , the charging station has to obtain @xmath91 and @xmath93 ( definitions [ defn : vlt ] & [ defn : v - k ] ) .
however , @xmath117 and @xmath93 do not depend on @xmath10 .
hence , we can separate the problem first the charging station finds @xmath91 and @xmath118 , and then it will select @xmath10 to fulfill the objective . _ we now focus on finding optimal _ @xmath10 .
first , we state the optimal values of the social welfare for any given realization of the user s utilites .
next , we state a pricing strategy which attains the above optimal value . note that if @xmath119 for each @xmath2,@xmath3 , and @xmath5 , then the social welfare is maximized when the user @xmath8 does not charge . in this case , the optimal value of social welfare is @xmath6 . on the other hand if @xmath120 for some @xmath2 , @xmath3 , and @xmath5 then the social welfare is maximized when the user @xmath8 charges its car .
if the user accepts the price menu @xmath10 , then the social welfare is @xmath121 .
thus , the maximum social welfare in the above scenario is @xmath122 as described in the following theorem [ thm : max_socialwelfare ] the maximum value of social welfare is @xmath123 . we obtain [ thm : max_socv2 g ] if @xmath124 , then it will maximize the ex - post social welfare . _ outline of the proof _ : if the price is set according to the above , the user will not select any contract if @xmath125 which gives @xmath6 social welfare .
the user select the contract which maximizes the payoff if @xmath126 . in this case , the ex - post social welfare is @xmath122 .
hence , the result follows
. however , the above pricing mechanism does not give any non - zero profit to the charging station .
the above pricing strategy is distribution independent , it holds for _ any arbitrary distribution .
_ also note that the pricing strategy also maximizes the social welfare in the long run when the additional cost of fulfilling a contract ( i.e. @xmath107 ) does not depend on the existing users in the charging station .
the condition that @xmath127 is independent of the existing evs in the charging station is satisfied if either all demand can be fulfilled using renewable energy or there is no renewable energy generation .
hence , in the _ two above extreme cases , the myopic pricing strategy is also optimal in the long run_.
we have already seen a pricing strategy which maximizes the ex - post social welfare in theorem [ thm : max_socv2 g ] , however , this pricing strategy does not give any positive profit .
naturally the question arises what is the pricing strategy that maximizes the expected profit of the charging station which will also maximize the ex - post social welfare .
we show that there exists a pricing strategy which may provide better profit to the charging station while maximizing the ex - post social welfare .
first , we introduce a notation which we use throughout .
[ defn : lower_endpoint ] let @xmath128 be the lowest end - point of the marginal distribution of the utility @xmath66 .
[ thm : profitmax_uncertainty ] consider the pricing strategy : @xmath129 the pricing strategy maximizes the ex - post social welfare .
the profit is @xmath130 .
_ outline of proof _ : first , note that adding a constant does not change the optimal solution .
hence , if @xmath131 , then @xmath132 is also optimal for price strategy in ( [ eq : pricesocialmax ] ) .
now , if the price is set according to ( [ eq : pricesocialmax ] ) , when @xmath133 , the user always select the contract .
if the condition is not satisfied , the user does not select the contract .
hence , the ex - post social welfare is always maximized .
note that if @xmath134 , then such a pricing strategy will provide a positive profit to the charging station .
hence , if @xmath107 is smaller than the lowest end - point of the distribution function , then the profit will be positive .
also note that the users which have higher utilities i.e. , higher @xmath128 will give more profits to the charging station .
the charging station needs to know the lowest end - points of the support set of the utilities unlike in theorem [ thm : max_socv2 g ] .
however , the charging station does not need to know the exact distribution functions of the utilities similar to theorem [ thm : max_socv2 g ] .
the lowest end - point can be easily obtained from the historical data .
the pricing strategy maximizes the ex - post social welfare similar to theorem [ thm : max_socv2 g ] .
this is also the _ maximum possible profit that the charging station can have under the condition that it maximizes the ex - post social welfare with probability @xmath135_. however , it may not maximize the expected profit of the charging station . in other words , the pricing strategy which maximizes the expected profit needs not maximize the ex - post social welfare .
* when the charging station is clairvoyant * : we have seen that if the charging station is unaware of the realized values of the utilities , there is no pricing strategy which maximizes both the expected profit and the ex - post social welfare .
however , we show that if the charging station is clairvoyant _
i.e. _ , it is aware of the realized values of the utilities of the users , then there exists a pricing strategy which both maximizes the social welfare and the profit of the charging station .
first , we introduce a notation .
[ defn : optimal ] let @xmath136 .
[ thm : profit_max ] let @xmath137 where @xmath138 is given in definition [ defn : optimal ] .
such a pricing strategy maximizes the profit as well as the social welfare .
_ there can be other pricing strategies which simultaneously maximize the social welfare and the profit .
_ though the joint profit and social welfare maximizing pricing strategy may not be unique , the profit of the charging station is the unique and is given by @xmath139 the above pricing strategy is an example of _ value - based _ pricing strategy where prices are set depending on the valuation or the utility of the users @xcite . in contrast
, the price strategy stated in theorem [ thm : max_socv2 g ] is an example of _ cost - based _ pricing strategy where the prices only depend on the costs . in the value - based pricing strategy ,
the user surplus decreases , in fact it is amount . in that case
, it will be @xmath140 optimal profit maximizing strategy . ]
@xmath6 in our case .
thus all the user surplus is transferred as the profit of the charging station .
_ thus , uncertainty regarding the utilities enhance the user s surplus .
_ theorem [ thm : profitmax_uncertainty ] entails that the charging station only has a positive profit if @xmath141 . if the above condition is not satisfied , the charging station s profit will be @xmath6 . in the following we consider a pricing strategy which will give a guaranteed positive profit to the charging station .
consider the pricing strategy @xmath142 where @xmath143 .
note that the pricing strategy stated in ( [ eq : pricesocialmax ] ) is a variant of the pricing strategy stated in ( [ eq : guranteed_profit ] ) where @xmath144 if we allow @xmath7 can also be @xmath6 .
the pricing strategy stated in ( [ eq : guranteed_profit ] ) gives the same positive profit irrespective of the menu selected by the user .
the regulator such as ferc can select a @xmath7 judiciously to trade off between the profit of the charging station and the social welfare . from lemma [ lm : v2 g ] , the pricing strategy stated in ( [ eq : guranteed_profit ] ) selects _ lower price for higher battery utilization .
_ this is desirable , as the user needs to be given incentive to battery utilization .
higher @xmath7 will deter the user s surplus . also note that when @xmath143 , it may not maximize the ex - post social welfare from theorem [ thm : profitmax_uncertainty ] .
very high value of @xmath7 also decreases the profit of the charging station , as users will be reluctant to accept any of the menus . the expected profit of the charging station for the above pricing strategy is [ thm : expected_payoff ] the expected profit of the charging station when it selects price according to ( [ eq : guranteed_profit ] ) is @xmath145
. _ outline of the proof _ : note that if a user selects any of the contracts , then the charging station s profit is @xmath7 .
hence , the charging station s expected profit is @xmath7 times the probability that at least one of the contracts will be accepted .
now , we provide an example where the pricing strategy in ( [ eq : guranteed_profit ] ) can also maximize the expected profit for a suitable choice of @xmath7 . first , we introduce a notation [ defn : alpha ] let @xmath146 . note that since @xmath66 is bounded and the probability distribution is continuous , thus , @xmath147 exists .
note from theorem [ thm : expected_payoff ] that @xmath147 corresponds to @xmath7 the charging station can get the maximum possible expected profit when the prices are of the form ( [ eq : guranteed_profit ] ) . if the utilities are drawn from a strictly increasing continuous distribution , then the set of @xmath148 would be singleton and we do not need to specify the maximum .
now , consider the pricing strategy @xmath149 where @xmath147 is as given in definition [ defn : alpha ] .
the above pricing strategy maximizes the profit for a class of utility functions which we describe below .
[ assum : utility ] suppose that the utility function @xmath150 for all @xmath2 , @xmath3 & @xmath5 ; @xmath151 is a constant and known to the charging station , and @xmath152 is a random variable whose realized value is not known to the charging station . in the above class of utility function
, the uncertainty is only regarding the realized value of the random variable @xmath152 .
note that @xmath152 is independent of @xmath2 , @xmath5 and @xmath3 , hence,@xmath152 is considered to be an additive noise .
it is important to note that we do not put any assumption whether _
@xmath152 should be drawn from a continuous or discrete distribution .
_ however , if the distribution is discrete , we need the condition that @xmath147 must exist .
[ thm : aclassutility ] the pricing strategy stated in ( [ eq : price_alpha ] ) maximizes the expected profit of the charging station ( given in ( [ eq : profitmax_expec ] ) ) when the utility functions are of the form given in assumption [ assum : utility ] . _
the above result is surprising .
it shows that a simple pricing mechanism such as the fixed profit can maximize the expected payoff for a large class of utility functions .
_ however , if the utilities do not satisfy assumption [ assum : utility ] the above pricing strategy may not be optimal .
the v2 g service will proliferate only if the users participate in that service .
the charging station can attain extra profits through the v2 g service .
however , the users will only select the menu with positive battery utilization if they get enough compensation .
thus , the charging station s profit inherently depends on whether the users have incentives to participate in the v2 g services . in this section
, we will analyze the conditions under which the users will be willing to participate in the v2 g service , and the profit of the charging station will increase .
first , we discuss the cost of battery utilization .
users will strictly prefer lower utilization as lower @xmath5 will increase the battery life . a higher battery utilization may increase the battery degradation cost @xcite .
we denote the cost associated with the utilization @xmath5 for user @xmath8 is @xmath153 where @xmath154 is a strictly increasing function .
the cost @xmath154 depends on the the state of the battery @xcite .
we assume that the user s utility @xmath66 is @xmath155 we consider that the cost function @xmath153 for the battery utilization as a linear function i.e. @xmath156 .
recently , @xcite shows that the per unit degradation cost for discharging remains almost constant for a wide range of values .
hence , a linear cost model can be a good approximation of the cost function .
however , our analysis can be easily extended to other cost models .
the charging station and even the user may not know the exact value of @xmath157 .
but , the ev manufacturer can easily provide the _
pessimistic _ approximation of @xmath157 such as the worst possible battery degradation cost for per unit of energy .
cents and @xmath158 cents . in this example @xmath157 can be taken as @xmath158 cents per kwh . ]
the realized value of the utility function of user @xmath8 is now @xmath159 .
note that if @xmath104 provides a positive payoff and a higher payoff to the user compared to @xmath160 , then the user will opt for v2 g service .
the following result formalizes the condition .
[ thm : condn ] user @xmath8 opts for battery degradation @xmath104 , if @xmath161 for some @xmath104 .
now , we consider the pricing strategy stated in ( [ eq : guranteed_profit ] ) i.e. , @xmath162 . note that with a linear _ battery degradation cost _ discussed in the last section @xmath163 our next result characterizes the condition for user s participation in the v2 g service for a linear battery degradation cost . from theorem [ thm : condn ]
if the following is satisfied : @xmath164 for some @xmath104,@xmath2,and @xmath3 ; then the user @xmath8 will have any incentive for v2 g service when the pricing strategy is as given in ( [ eq : guranteed_profit ] ) .
the expected profit of the charging station also increases under the above condition .
_ outline of the proof _ : note that @xmath165 .
thus , the user will never select the contract with @xmath160 if @xmath164 for some @xmath104 . in the fixed profit scheme ,
the charging station always gets a profit of @xmath7 for if a contract is selected . when @xmath164 for some @xmath104 , the user s probability of selecting any contract increases . hence , the expected profit of the charging station increases from theorem 5 .
thus , if @xmath164 for some @xmath104 , the user s surplus and the profit of the charging station both increase . if the harvested renewable energy is large enough , then the charging station can fulfill the demand using the renewable energy .
hence , the difference between @xmath91 and @xmath166 will be not enough for the user to participate in the v2 g service .
on the other hand , if the harvested renewable energy is small , then the charging station may have to buy expensive conventional energy from the grid to fulfill the demand .
hence , the user will have a higher incentive to participate in the v2 g service as the difference between @xmath91 and @xmath166 may be significant .
the difference is more significant when the conventional energy is more expensive ( e.g. peak period ) .
_ thus , the user will have a higher incentive to participate in the v2 g service when the renewable energy generation is small and the cost of the conventional energy is high . _ if the storage capacity of the charging station is large , the charging station may buy energy from the grid during the off - peak period and use it during the peak period .
this also reduces the difference between @xmath166 and @xmath91 .
however , if the storage capacity is low , the difference between @xmath166 and @xmath91 again increases .
hence , _ the user s participation towards the v2 g service is more likely when the storage capacity of the charging station is small and the renewable energy harvesting is low .
_ _ this shows the necessity of the v2 g service for better profitability as the high storage capacity is very costly to procure and the penetration of the renewable energy is still low .
_ so far , we assumed that evs only come for charging where @xmath167 for menu - pricing .
however , once the v2 g service proliferates , the users with their fully charged batteries may come during the peak hours to the charging station in order to _ only _ discharge
. the users can charge their batteries again in their homes during the off - peak times . in this manner
, the users can gain some profits .
though , we have not explicitly considered this scenario , our model can be easily extended to the above scenario .
for example , the price @xmath168 where @xmath169 will denote that the user @xmath8 s ev will be discharged _ at most _ @xmath2 amount within deadline @xmath3 and additional battery utilization @xmath170 .
the negative price indicates that the charging station will pay to the user , as the user is delivering energy to the grid .
note that the _ only _ constraint needs to be changed is to replace @xmath2 with @xmath171 in ( [ eq : battery_utilization ] ) . since @xmath172 still remains a linear programming problem ( definition [ defn : vlt ] ) , the charging station can still find @xmath107 and can select prices according to the strategies discussed before .
hence , our results also holds in this scenario .
we numerically study and compare various pricing strategies presented in this paper .
we evaluate the profit of the charging station and the user s surplus achieved in those pricing strategies .
we also analyze the impact of the v2 g service .
similar to @xcite , the user s utility for energy @xmath173 is taken to be of the form @xmath174 thus , the user s utility is a strictly increasing and concave function in the amount energy consumed @xmath173 . the quadratic utility functions for ev charging have also been considered in @xcite .
note that the user s desired level of charging is @xmath175 .
_ we assume that @xmath175 is a random variable_. @xcite shows that in a commercial charging station , the average amount of energy consumed per ev is @xmath176kwh with standard deviation @xmath177kwh .
we assume that @xmath175 is a truncated gaussian random variable with mean @xmath176kwh and standard deviation @xmath177kwh , where the truncation is to the interval @xmath178 $ ] .
we assume that the maximum battery capacity is @xmath179 , and the minimum capacity as @xmath180 .
the initial battery level of a new user is assumed to be uniformly distributed in the interval @xmath181 $ ] .
note that the upper bound is @xmath182 , since the user s desired level of charging is @xmath175 .
following @xcite , the deadline or the time spent by an electric vehicle in a commercial charging is distributed with an exponential distribution with mean @xmath183 hours .
thus , the preferred deadline ( @xmath184 ) of the user to be an exponentially distributed random variable with mean @xmath183 . users strictly prefer a lower deadline for a given energy .
we assume the utility as a convex decreasing function of the deadline .
the utility of the user after the preferred deadline is assumed to be @xmath6 .
hence , the user s utility for energy @xmath2 and deadline @xmath3 is chosen as the following @xmath185 the cost for the battery utilization is taken as @xmath186 .
we assume that @xmath187 .
thus , the total utility for the user for contract @xmath0 is @xmath188 the arrival process of electric vehicles is assumed to be a non - homogeneous poisson arrival process since the arrival rates vary over time .
for example , during the peak - hours ( 8 am to 5 pm ) the arrival rate is higher compared to the off - peak hours .
the arrival rates are chosen as @xmath189 ( @xmath190 , resp . )
vehicles per hour during the peak period ( off - peak period , resp . ) .
we also assume that the maximum charging rate @xmath39 is @xmath191 kw .
we also assume @xmath40 as @xmath192 kw . the energy harvesting device is assumed to harvest energy which is a truncated gaussian random variable with mean @xmath193 and standard deviation @xmath135 per hour .
initial battery level is assumed to be @xmath6 _ i.e. _ it is fully discharged .
the prices @xmath81 for the conventional energy is assumed to be governed by time - of - use ( tou ) time scale .
thus , the cost of buying conventional energy varies over time .
we assume that the selling price , @xmath194 .
we assume that the charging station is unaware of the exact utilities .
however , it knows the lower endpoints of the support sets .
the charging station also knows the value of @xmath195 and the structure of the utility function , but does not know the realized values .
we consider the pricing strategy that we have introduced in section [ sec : price_uncertainty ] , which is given here @xmath196 recall from definition [ defn : lower_endpoint ] that @xmath197 is the lowest end - point of the utility @xmath198 we study the impact of @xmath7 .
we also study the impact of the maximum storage capacity of the charging station @xmath41 . towards this end , we consider two different values of @xmath41 ; a low value , @xmath199kwh , and a high value , @xmath200kwh . fig .
[ fig : betavsadmittedusers ] shows that as @xmath7 increases the number of admitted users decreases .
however , the decrement is slow initially .
when @xmath7 becomes larger than a threshold , the price selected to the users becomes very large , and thus , fewer number of evs are admitted . note
that v2 g service enables the charging station to admit more users .
[ fig : consumer_surplus ] shows that as @xmath7 increases , the user s surplus decreases , as the users pay more .
the decrement is not significant for small values of @xmath7 .
also , note that v2 g enhances the user s surplus even though the users incur costs for the v2 g service .
as @xmath7 increases the profit increases initially .
however , when @xmath201 , the number of users served decreases rapidly , hence , the profit also drops .
v2 g service significantly increases the profit of the charging station . at high values of @xmath7
both users surpluses and the profit decrease significantly . low values of @xmath7 gives high users surpluses , however , the profit is low .
@xmath202 $ ] is the best candidate for the balance between profit and users surpluses .
[ fig : soc_wellowb ] shows that if the charging station s storage capacity is low , the user s surplus and the profit decrease .
this is because when the storage capacity is low , the charging station can only store lower amount of energy for the future use .
hence , the charging station has to pay more to buy conventional energy during the peak period .
hence , the charging station s profit and the user s surplus both decrease .
we subsequently study the effect of @xmath7 on the time spent by the users in the charging station ( fig .
[ fig : deadline ] ) .
our analysis shows that users spend more time in the charging station with the increase in @xmath7 . as @xmath7 increases
, the users which have preferences for lower deadlines have to pay more , since the cost of fulfilling lower deadline contracts is high . hence , those users are reluctant to accept the contract .
thus , the accepted users spend more time in the charging station .
[ fig : deadline ] also shows that the evs spend lower amount of time when @xmath41 is high .
this is because when @xmath41 is high , the charging station s cost to fulfill contracts with a lower deadline is lower compared to low @xmath41 .
[ fig : bu ] shows the average amount of battery utilization of each ev as a function of @xmath7 . as @xmath7 increases the average battery utilization increases .
when @xmath7 is high , the user has to pay a larger amount .
however , if the user agrees for higher battery utilization , then the price may be low because the charging station may incur a lower cost .
hence , the average battery utilization increases as @xmath7 becomes high .
since the charging station can only store a lower amount of renewable energy and a lower amount of energy bought from the grid during the off - peak periods , hence , the average battery utilization is higher for lower @xmath41 .
the batteries of evs are used more for lower storage capacity .
[ fig : active_max ] shows the impact of @xmath7 on the _ maximum _ number of evs that can simultaneously charge ( or , the number of active users ) along time .
this given an idea of how much charging spots we need , in order to implement the pricing strategy .
since the average time spent by users in the charging station increases with @xmath7 and the number of admitted users are almost the same for @xmath203 , the number of active users increases as @xmath7 increases .
since when @xmath41 is low the average time spent by an ev increases , thus , the number of active users is higher when @xmath41 is low .
[ fig : conven ] shows the energy bought from the grid over the day .
the figure shows that the peak is significantly reduced during the on - peak period .
in fact , the peak is shifted to the off - peak period even though more users come during the on peak period .
this is because during the on - peak periods , the evs are mainly used for discharging or the charging station uses more energy from its storage . during the off - peak period ,
the evs are charged , and energy is stored in the battery of the charging station .
[ fig : diffbuy ] shows that the peak energy used from the grid is higher when @xmath41 is high .
this is because when @xmath41 is high , more energy is bought and stored during the off - peak period which results in a higher peak .
the reduction of the peak decreases very slowly with @xmath7 initially
. however , as @xmath7 becomes very high , very few users are admitted which results in the drop of peak energy .
as @xmath7 increases the users with higher utilities will accept the contracts .
thus , users with higher energy demand will be more likely to accept the contract and hence , the average charging amount for each ev should increase with @xmath7 . however , fig .
[ fig : energy_mean ] shows that for @xmath203 , the average energy consumed by each ev does not increase significantly with the increase in @xmath7 .
the apparent anomaly is due to the fact that the users with higher demand but with smaller deadline preferences , may have to pay more because of the increase in the price to fulfill the contract as @xmath7 increases .
hence , such users will not accept the offers which results in a decrease of the average energy consumption with the increase in @xmath7 initially .
however , as @xmath7 increases only the users with higher demand accept offers increasing the average energy consumption .
since the user spends more time for lower @xmath41 , the average energy consumed by each ev is slightly higher for low @xmath41 .
[ fig : price ] shows that as @xmath7 increases the price increases .
interestingly , price is not higher during the peak - period , since more evs are involved in the v2 g service during the peak period . during the off - peak period
, the price from the conventional energy is low which does not provide any incentive for the v2 g service .
a new price is selected every - time an ev is admitted .
however , our study shows that the variation of prices within a time interval is nominal .
this paper proposed a menu - based pricing approach for v2 g service where the charging station selects prices to the arriving users for different charging amounts , additional amount of battery utilization , and the deadline .
the user either selects one of the contracts or rejects all based on her utility function .
when the charging station is aware of the utilities , there exists a profit maximizing pricing strategy which can also maximize the social welfare , however , it provides zero user s surplus .
however , when the charging station is not aware of the utilities there is _ no _ profit maximizer pricing strategy which can maximize the ex - post social welfare .
this paper also considered a pricing strategy which gives a fixed positive profit to the charging station irrespective of the contract chosen .
the conditions where the users have incentives to participate in the v2 g service are characterized , and it is seen that the v2 g service is more preferable when the harvested energy is low , and the storage capacity of the charging station is low .
this work can be extended in several directions .
for example , the characterization of the optimal price in the non - myopic setting remain open .
this paper considered a fixed price for selling the energy to the grid .
the optimal price that the grid should set to enhance the v2 g service or reduce the peak consumption constitutes a future research direction .
e. sortomme , m. m. hindi , s. d. j. macpherson , and s. s. venkata , `` coordinated charging of plug - in hybrid electric vehicles to minimize distribution system losses , '' _ ieee transactions on smart grid _ ,
vol . 2 , no . 1 , pp .
198205 , march 2011 .
m. alizadeh , h. t. wai , m. chowdhury , a. goldsmith , a. scaglione , and t. javidi , `` optimal pricing to manage electric vehicles in coupled power and transportation networks , '' _ ieee transactions on control of network systems _ , vol .
pp , no .
99 , pp . 11 , 2016 .
j. xu and v. w. s. wong , `` an approximate dynamic programming approach for coordinated charging control at vehicle - to - grid aggregator , '' in _ smart grid communications ( smartgridcomm ) , 2011 ieee international conference on _ , oct 2011 , pp . 279284 .
h. xing , m. fu , z. lin , and y. mou , `` decentralized optimal scheduling for charging and discharging of plug - in electric vehicles in smart grids , '' _ ieee transactions on power systems _ , vol .
31 , no . 5 , pp . 41184127 , sept 2016 . b. g. kim , s. ren , m. van der schaar , and j. w. lee , `` bidirectional energy trading and residential load scheduling with electric vehicles in the smart grid , '' _ ieee journal on selected areas in communications _ , vol . 31 , no . 7 , pp .
12191234 , july 2013 .
a. nayyar , m. negrete - pincetic , k. poolla , and p. varaiya , `` duration - differentiated energy services with a continuum of loads , '' _ corr _ , vol .
abs/1408.5825 , 2014 .
[ online ] .
available : http://arxiv.org/abs/1408.5825 a.ghosh and v.aggarwal , `` control of charging of electric vehicles through menu based pricing , '' 2016 .
[ online ] .
available : http://arxiv.org/abs/1609.09037 m. a. delucchi and t. e. lipman , `` an analysis of the retail and lifecycle cost of battery - powered electric vehicles , '' _ transportation research part d : transport and environment _ , vol .
6 , no . 6 , pp . 371 404 , 2001 . c. zhou , k. qian , m. allan , and w. zhou , `` modeling of the cost of ev battery wear due to v2 g application in power systems , '' _ ieee transactions on energy conversion _ ,
26 , no . 4 , pp .
10411050 , dec 2011 .
w. shuai , p. maill , and a. pelov , `` charging electric vehicles in the smart city : a survey of economy - driven approaches , '' _ corr _ , vol .
abs/1601.03925 , 2016 .
[ online ] .
available : http://arxiv.org/abs/1601.03925 l.
gan , u. topcu , and s. h. low , `` optimal decentralized protocol for electric vehicle charging , '' _ ieee transactions on power systems _ , vol .
28 , no . 2 , pp .
940951 , may 2013 . `` evaluating electric vehicle charging impacts and customer charging behaviors- experiences from six smart grid investment grant projects , '' https://www.smartgrid.gov/files/b3_revised_master-12-17-2014_report.pdf , december , 2014 .
now , since @xmath209 , and @xmath210 , thus , @xmath211 .
thus , @xmath212 .
hence , @xmath213 the other constraints are independent of @xmath214 , hence , it will be feasible .
thus , with the modified @xmath215 , @xmath216 , and @xmath217 , with the other unchanged solution the value of the optimization problem is _
lower_. hence , this leads to a contradiction .
now , we show that @xmath28 in an optimal solution of ( [ eq : vlt ] ) .
suppose that in an optimal solution we have @xmath218 and @xmath219 for some @xmath29 and @xmath3 .
now , consider the following modification for each @xmath29 .
@xmath220 note that because of the above modification @xmath221 hence , the constraints in ( [ eq : charged ] ) , ( [ eq : chargeeachev ] ) , ( [ eq : dischargelimit ] ) , ( [ eq : limit ] ) , ( [ eq : battery_utilization ] ) are satisfied for this modified solution .
now , consider the following @xmath222 since @xmath223 , and @xmath224 for all @xmath29 and @xmath225 for at least index @xmath30 , thus , @xmath226 .
thus , if @xmath227 and @xmath228 are modified as follows @xmath229 then the constraints in ( [ eq : battery_capacity ] ) , ( [ eq : battery_level ] ) are also satisfied when the rest of the solution are unchanged .
however , the cost decreases as @xmath230 with the modified values of @xmath231 , @xmath232 , @xmath233 .
this contradicts that @xmath234 are optimal .
hence , the result follows .
the above theorem also shows the transformation of the optimal solution requires to get @xmath53 ( cf.([eq : transformse ] ) , ( [ eq : lowervaluee ] ) ) and @xmath28 ( cf.([eq : transformr ] ) , ( [ eq : transformrtch ] ) , ( [ eq : lowervaluer ] ) ) in case if @xmath235 are bounded
. the modified solution will still be bounded .
suppose the statement is false .
without loss of generality , assume that @xmath236 where @xmath237 for some @xmath5 , @xmath2 and @xmath3 achieves a strictly higher expected payoff than the pricing strategy @xmath238 .
the expected profit of the charging station for pricing strategy @xmath236 is given by @xmath239 now , we evaluate the expression @xmath240 .
the user @xmath8 will select the menu @xmath10 with a positive probability if @xmath241 and for every @xmath242 , @xmath243 since @xmath244 are fixed , hence , the above inequality is either satisfied or not satisfied with probability @xmath135 .
more specifically , the user selects the menu @xmath0 if @xmath241 and @xmath245 without loss of generality , assume that @xmath246 be the maximum value for which the above inequality is satisfied i.e. @xmath247 the random variable @xmath152 only affects the probability whether @xmath248 or not .
hence , the charging station s expected profit is upper bounded by @xmath249 note that by the definition of @xmath147 ( definition [ defn : alpha ] ) , @xmath250 however by theorem [ thm : expected_payoff ] the expected payoff of the charging station when it selects the price @xmath251 is given by the expression in the left - hand side of the inequality in ( [ eq : compare ] ) .
hence , this leads to a contradiction .
thus , the result follows . | the paper considers a bidirectional power flow model of the electric vehicles ( evs ) in a charging station .
the evs can inject energies by discharging via a vehicle - to - grid ( v2 g ) service which can enhance the profits of the charging station .
however , frequent charging and discharging degrade battery life .
a proper compensation needs to be paid to the users to participate in the v2 g service .
we propose a menu - based pricing scheme , where the charging station selects a price for each arriving user for the amount of battery utilization , the total energy , and the time ( deadline ) that the ev will stay .
the user can accept one of the contracts or rejects all depending on their utilities .
the charging station can serve users using a combination of the renewable energy and the conventional energy bought from the grid . we show that though there exists a profit maximizing price which maximizes the social welfare , it provides no surplus to the users if the charging station is aware of the utilities of the users .
if the charging station is not aware of the exact utilities , the social welfare maximizing price may not maximize the expected profit .
in fact , it can give a zero profit .
we propose a pricing strategy which provides a guaranteed fixed profit to the charging station and it also maximizes the expected profit for a wide range of utility functions .
our analysis shows that when the harvested renewable energy is small the users have higher incentives for the v2 g service .
we , numerically , show that the charging station s profit and the user s surplus both increase as v2 g service is efficiently utilized by the pricing mechanism . |
let @xmath1 be an algebraic variety in @xmath2 .
the first section discusses the _ logarithmic limit set _ @xmath3 of @xmath1 , which we use to define the set of ideal points of @xmath1 .
we show that each point with rational coordinate ratios in @xmath3 is an ideal point of a curve in @xmath1 .
thus , the logarithmic limit set lends itself to applications of culler shalen theory , where essential surfaces are associated to ideal points of curves in the character variety of a 3manifold . given a compact , orientable , irreducible 3manifold with boundary consisting of a single torus ,
one calls boundary slopes of essential surfaces associated to ideal points of the character variety _ strongly detected_. it is shown in @xcite that the slope of each side of the newton polygon of the @xmath0polynomial is a strongly detected boundary slope . in the second section ,
we generalise the @xmath0polynomial to define an _
eigenvalue variety _ of a manifold with boundary consisting of @xmath4 tori .
we describe the _ strongly detected boundary curves _ , i.e. the boundary curves of essential surfaces associated to ideal points of the character variety , via a map from the logarithmic limit set of the eigenvalue variety to the @xmath5sphere .
the third section contains a computation of the @xmath0polynomials of torus knots .
these seem to be the only infinite families of @xmath0polynomials currently known .
let @xmath1 be a subvariety of @xmath6 defined by an ideal @xmath7 , and let @xmath8=\c [ x_1^{\pm 1 } , \ldots , x_m^{\pm 1}]$ ] .
bergman gives the following three descriptions of a _ logarithmic limit set _ in @xcite : \1 .
define @xmath9 as the set of limit points on @xmath10 of the set of real @xmath11tuples in the interior of @xmath12 : @xmath13 \2 .
define @xmath14 as the set of @xmath11tuples @xmath15 as @xmath16 runs over all real valued valuations on @xmath8/j$ ] satisfying @xmath17 .
define the support @xmath18 of an element @xmath19 $ ] to be the set of points @xmath20 such that @xmath21 occurs with non
zero coefficient in @xmath22 . then define @xmath23 to be the set of @xmath24 such that for all non
zero @xmath25 , the maximum value of the dot product @xmath26 as @xmath27 runs over @xmath18 is assumed at least twice .
bergman shows in @xcite that the second and third descriptions are equivalent , and it follows from bieri and groves @xcite that all descriptions are equivalent : @xmath28 .
we therefore write @xmath3 for the logarithmic limit set of the variety @xmath1 . a generic picture for the projection of a variety in @xmath29 into @xmath30 and its logarithmic limit set in @xmath31
is given in figure [ fig : loglim ] .
the ideal generated by a set of polynomials @xmath32 shall be denoted by @xmath33 , and the variety generated by the ideal @xmath34 is denoted by @xmath35 or @xmath36 , if @xmath37 . , width=188 ] the third description of the logarithmic limit set can be illustrated using spherical duals of newton polytopes .
let @xmath38 $ ] be given as an expression @xmath39 .
then the _ newton polytope _ of @xmath22 is the convex hull of @xmath40 .
thus , if @xmath41 , then @xmath42 note that @xmath43 , and since @xmath44 , we have @xmath45 . for arbitrary subsets @xmath46 of @xmath47 ,
let @xmath48 .
the following facts about newton polytopes of polynomials are well known ( cf .
@xcite ) : 1 . let @xmath49 $ ] such that @xmath50
. then @xmath51 .
2 . @xmath52 .
the _ spherical dual _ of a bounded convex polytope @xmath53 in @xmath47 is the set of vectors @xmath54 of length @xmath55 such that the supremum @xmath56 is achieved for more than one @xmath27 , and it is denoted by @xmath57 .
geometrically , @xmath58 consists of all outward pointing unit normal vectors to the support planes of @xmath53 which meet @xmath53 in more than one point .
if @xmath53 is the newton polytope of a non zero polynomial @xmath22 , then the spherical dual of @xmath59 is also denoted by @xmath58 .
the following lemma is immediate from the third description of the logarithmic limit set : [ log : int lem ] let @xmath1 be an algebraic variety in @xmath60 defined by the ideal @xmath7 . then @xmath3 is the intersection over all non zero elements of @xmath7 of the spherical duals of their newton polytopes .
spherical duals are easy to visualise in low dimensions . given a convex polygon @xmath53 in @xmath61 , its spherical dual is the collection of points on the unit circle defined by outward pointing unit normal vectors to edges of @xmath53 . given a convex polyhedron in @xmath62 , we obtain the vertices of its spherical dual again as points on the unit sphere @xmath63 arising from outward pointing unit normal vectors to faces .
we then join two of these points along the shorter geodesic arc in @xmath63 if the corresponding faces have a common edge .
this gives a finite graph in @xmath63 .
[ [ section ] ] if the ideal @xmath7 is principal , defined by an element @xmath64 , then bergman states in @xcite that the set @xmath65 is precisely @xmath58 .
we give a proof using the following lemma .
[ log : sph lem ] let @xmath49 $ ] .
if @xmath50 , then @xmath66 .
if @xmath67 , then there are distinct @xmath68 such that @xmath69 since @xmath70 is convex and bounded , there is @xmath71 such that @xmath72 since @xmath51 , it follows that @xmath73 and @xmath74 thus , @xmath75 , and by symmetry @xmath76 .
now then assume that @xmath77 .
then there are distinct @xmath78 and @xmath79 in @xmath80 such that @xmath81 furthermore , there are @xmath82 and @xmath83 such that @xmath84 . if @xmath85 , then @xmath86 and @xmath87 , contradicting the choice of @xmath78 .
thus , @xmath88 , and similarly @xmath89 .
a similar argument shows : @xmath90 and similarly for the @xmath91 . since @xmath92 , either @xmath93 or @xmath94 , and hence either @xmath67 or @xmath95 .
let @xmath38 $ ] .
if @xmath64 , then @xmath96 .
since each non zero element of @xmath97 is of the form @xmath98 for some non
zero @xmath99 $ ] , lemma [ log : int lem ] and lemma [ log : sph lem ] yield : @xmath100}{\bigcap } sph(fg ) = \underset{0 \ne g \in \sc[x^\pm ] } { \bigcap } ( sph(f ) \cup sph(g ) ) = sph(f),\ ] ] since @xmath101 .
logarithmic limit sets of varieties defined by a principal ideal are therefore completely determined by a generator of the ideal , and readily computable .
if the ideal @xmath7 is not principal , then @xmath102 is the intersection of the spherical duals of the newton polytopes of _ finitely _ many elements in @xmath7 .
this follows from bieri and groves @xcite .
however , there is presently no known algorithm to determine a suitable finite set of elements of @xmath7 .
conjecturally , a grbner basis may suffice .
[ [ log : computing ] ] as bergman notes , the spherical dual of the convex hull of a finite subset @xmath103 of cardinality @xmath104 is a finite union of convex spherical polytopes .
it is the union over all @xmath105 of the set of @xmath54 satisfying the @xmath106 inequalities @xmath107 where @xmath27 ranges over @xmath0 .
this may be useful for calculations , but it also shows that corners of the convex spherical polytopes which we consider have rational coordinate ratios , and points with rational coordinate ratios are dense in the polytope .
such a polytope is called a _ rational convex spherical polytope_. we summarise further results by bergman and bieri groves as follows : [ bergman - bieri - groves ] the logarithmic limit set @xmath3 of an algebraic variety @xmath1 in @xmath6 is a finite union of rational convex spherical polytopes .
the maximal dimension for a polytope in this union is achieved by at least one member @xmath53 in this union , and we have @xmath108 bergman asks in @xcite whether every polytope in the above union always has the same dimension . in @xcite
it is shown that the irreducible component in the @xmath109eigenvalue variety of the whitehead link complement @xmath110 which arises from the dehn surgery component has logarithmic limit set consisting of five components , four of which are points , and one of which is a union of geodesic arcs .
a schematic picture is given in figure [ fig : whl_log_lim ] .
it would be interesting to modify bergman s question and ask whether each connected component of the logarithmic limit set of an irreducible variety can always be expressed as a union of polytopes all having the same dimension when @xmath111 .
, width=188 ] we conclude this section with a lemma , which is useful for applications of culler shalen theory .
[ curve finding lemma ] let @xmath1 be an algebraic variety in @xmath6 , and @xmath112 be a point with rational coordinate ratios .
then there is a curve @xmath113 , i.e. an irreducible subvariety of complex dimension one , in @xmath1 such that @xmath114 . by a lemma of bergman ( @xcite , p.464 ) , the point @xmath54 is contained in the logarithmic limit set of an irreducible component of @xmath1 .
thus , we may assume that @xmath1 is irreducible for the purpose of this proof .
assume that @xmath1 is defined by the ideal @xmath34 in @xmath115 = \c[x_1^{\pm 1},\ldots , x_m^{\pm 1}]$ ] .
we follow word by word bergman s construction starting after `` ( 4 ) '' on page 465 , up to the sentence `` hence the points @xmath116 will approach @xmath117 . '' on page 466 . here
@xmath118 denotes @xmath119 .
our adaption reads as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ let @xmath54 be a point of @xmath3 the ratios of whose coordinate functions are rational .
making use of the action of @xmath120 on @xmath115 $ ] and the induced action on @xmath10 ( not isometric ! ) , we can reduce to the case where @xmath121 .
let @xmath122 designate the ring @xmath123 \subset r = \c[x^{\pm}]$ ] , and @xmath124 the ring @xmath125 $ ] , into which we map @xmath122 by sending @xmath126 to @xmath127 .
let @xmath128 designate @xmath129 , and @xmath130 the image of @xmath128 in @xmath124 .
thus , @xmath130 consists of the components of degree @xmath127 in @xmath126 of elements of @xmath128 .
let @xmath131 designate the subvariety of @xmath132 defined by @xmath128 , and @xmath133 , the subvariety of @xmath134 defined by @xmath130 .
@xmath131 will be the zariski closure of @xmath1 in @xmath132 , and @xmath135 .
if @xmath136 , this means that every nonzero @xmath137 has more than one term of minimal degree in @xmath126 .
hence every nonzero element of @xmath130 has more than one term , i.e. @xmath130 is a proper ideal of @xmath124 .
then @xmath138 is non empty : let @xmath139 .
now @xmath1 will in fact be dense in @xmath131 under the topology induced by the absolute value on @xmath140 .
[ ] hence we can obtain @xmath141 as a limit of points @xmath142 .
as @xmath143 approaches @xmath141 , @xmath144 will approach the finite values @xmath145 for @xmath146 , but @xmath147 will approach @xmath148 . hence the points @xmath116 will approach @xmath149 . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we construct a curve @xmath113 in @xmath1 such that @xmath150 using the following two facts from @xcite : 1 . ( 1.14 ) let @xmath151 be an irreducible affine variety . if @xmath152 is a proper subvariety of @xmath151 , then @xmath153 .
2 . ( 3.14 ) if @xmath152 is an r dimensional affine variety and @xmath154 are polynomial functions on @xmath152 , then every component of @xmath155 has dimension @xmath156 .
since @xmath138 is a proper subvariety of @xmath131 , obtained by intersection with a hyperplane , we have @xmath157 , and since @xmath158 , we have @xmath159 . we need to find a curve @xmath160 in @xmath131 which has non trivial intersection with both @xmath138 and @xmath1 , since then @xmath161 is a curve in @xmath1 such that @xmath162 .
let @xmath163 .
let @xmath139 and choose a polynomial @xmath164 on @xmath138 such that @xmath165 , but @xmath164 is not identical to @xmath127 on @xmath138 .
then @xmath164 is not identical to @xmath127 on @xmath131 and each component of the variety @xmath166 has dimension exactly @xmath167 , and intersects @xmath138 in a @xmath168dimensional subvariety .
let @xmath169 be a component of @xmath166 containing @xmath141 .
choose a polynomial @xmath170 such that @xmath171 , but @xmath170 is not identical to @xmath127 on @xmath172 .
then each component of @xmath173 has dimension exactly @xmath174 and intersects @xmath138 in a @xmath175dimensional subvariety .
let @xmath176 be a component of @xmath177 containing @xmath141 .
we now proceed inductively for @xmath178 .
given a component @xmath179 of @xmath180 which is a @xmath181dimensional subvariety of @xmath131 , with the property that it contains @xmath141 , but also points of @xmath182 , we let @xmath183 be a polynomial function on @xmath179 which vanishes on @xmath141 but not on all of @xmath184 . then each component of @xmath185 has dimension exactly @xmath186 and intersects @xmath138 in a @xmath187dimensional subvariety , and we let @xmath188 be a component of @xmath185 containing @xmath141 .
then @xmath189 is a curve in @xmath1 with the desired property .
a _ surface _ @xmath190 in a compact 3manifold @xmath191 will always mean a 2dimensional pl submanifold _ properly embedded _ in @xmath191 , that is , a closed subset of @xmath191 with @xmath192 .
if @xmath191 is not compact , we replace it by a compact core .
an embedded sphere @xmath63 in a 3manifold @xmath191 is called _ incompressible _ if it does not bound an embedded ball in @xmath191 , and a manifold is _ irreducible _ if it contains no incompressible 2spheres .
an orientable surface @xmath190 without 2sphere or disc components in an orientable 3manifold @xmath191 is called _ irreducible _ if for each disc @xmath193 with @xmath194 there is a disc @xmath195 with @xmath196 .
we will also use the following definition : @xcite[def : essential ] a surface @xmath190 in a compact , irreducible , orientable 3manifold is said to be essential if it has the following five properties : 1 .
@xmath190 is bicollared ; 2 .
the inclusion homomorphism @xmath197 is injective for every component @xmath198 of @xmath190 ; 3 .
no component of @xmath190 is a 2sphere ; 4 .
no component of @xmath190 is boundary parallel ; 5 .
@xmath190 is nonempty .
the following definition and facts can be found in @xcite .
a ( properly embedded ) surface @xmath190 in a compact 3manifold @xmath191 is _
@xmath200incompressible _ if for each disc @xmath193 such that @xmath201 splits into two arcs @xmath27 and @xmath202 meeting only at their common endpoints with @xmath203 and @xmath204 there is a disc @xmath195 with @xmath205 and @xmath206 . a surface @xmath190 is incompressible and @xmath200incompressible if and only if each component of @xmath190 is incompressible and @xmath200incompressible . @xcite [ ev : bdry incomp ] let @xmath190 be a connected incompressible surface in the irreducible 3manifold @xmath191 , with @xmath207 contained in torus boundary components of @xmath191 .
then either @xmath190 is @xmath200incompressible or @xmath190 is a boundary parallel annulus . for the remainder of this section , we let @xmath191 be a compact , orientable , irreducible 3manifold with non empty boundary consisting of a disjoint union of @xmath4 tori .
it follows from definition [ def : essential ] and lemma [ ev : bdry incomp ] that an essential surface in @xmath191 is @xmath200incompressible .
denote the boundary tori of @xmath191 by @xmath208 , and choose an oriented meridian @xmath209 and an oriented longitude @xmath210 for each @xmath211 .
let @xmath190 be an incompressible and @xmath200-incompressible surface with non empty boundary in @xmath191 .
there are coprime integers @xmath212 such that @xmath190 meets @xmath211 in @xmath213 curves parallel to @xmath214 .
we thus obtain a point @xmath215 we view the above point as an element in @xmath216 since most methods of detecting essential surfaces only do so up to projectivisation .
furthermore , we ignore the orientation of @xmath207 , so the loop @xmath214 is equivalent to @xmath217 .
all we have to know are the relative signs of @xmath218 and @xmath219 .
this equivalence induces a @xmath220 action on @xmath216 , and the closure of the set of _ projectivised boundary curve coordinates _ : @xmath221 \in \rr p^{2h-1}/\z_2^{h-1}\ ] ] arising from incompressible and @xmath200incompressible surfaces with non empty boundary in @xmath191 is called the _ boundary curve space _ of @xmath191 and denoted by @xmath222 .
note that @xmath223 ( see @xcite ) .
we will use the following result .
@xcite [ ev : dim bc ] let @xmath191 be a compact , orientable , irreducible 3manifold with non empty boundary consisting of a disjoint union of @xmath4 tori
. then @xmath224 .
let @xmath225 be a finitely generated group .
the set of representations of @xmath225 into @xmath226 is @xmath227 .
this set is called the _
representation variety _ of @xmath225 , and regarded as an affine algebraic set .
each @xmath228 defines a _ trace function _ @xmath229 by @xmath230 , which is an element of the coordinate ring @xmath231 $ ] .
two representations are _ equivalent _ if they differ by an inner automorphism of @xmath226 .
a representation is _ irreducible _ if the only subspaces of @xmath232 invariant under its image are trivial .
this is equivalent to saying that the representation can not be conjugated to a representation by upper triangular matrices .
otherwise a representation is _
reducible_. for each @xmath233 , its _ character _ is the function @xmath234 defined by @xmath235 .
it turns out that irreducible representations are determined by characters up to equivalence , and that the reducible representations form a closed subset @xmath236 of @xmath237 .
the collection of characters @xmath238 can be regarded as an affine algebraic set , which is called the _ character variety_. there is a regular map @xmath239 taking representations to characters . if @xmath225 is the fundamental group of a topological space @xmath191 , we also write @xmath240 and @xmath241 instead of @xmath237 and @xmath238 respectively .
there also is a character variety arising from representations into @xmath109 , see boyer and zhang @xcite , and the relevant objects are denoted by placing a bar over the previous notation .
the natural map @xmath242 is finite to one , but in general not onto .
as with the @xmath226character variety , there is a surjective `` quotient '' map @xmath243 , which is constant on conjugacy classes , and with the property that if @xmath244 is an irreducible representation , then @xmath245 is the orbit of @xmath244 under conjugation .
again , a representation is irreducible if it is not conjugate to a representation by upper triangular matrices .
if a 3manifold @xmath191 admits a complete hyperbolic structure of finite volume , then there is a discrete and faithful representation @xmath246 .
this representation is neccessarily irreducible , as hyperbolic geometry otherwise implies that @xmath191 has infinite volume .
if @xmath191 is not compact , then a _ compact core _ of @xmath191 is a compact manifold @xmath247 such that @xmath191 is homeomorphic to the interior of @xmath247 .
we will rely on the following result : @xcite [ thurston thm ] let @xmath191 be a complete hyperbolic 3manifold of finite volume with @xmath4 cusps , and let @xmath248 be a discrete and faithful representation associated to the complete hyperbolic structure .
then @xmath249 admits a lift @xmath250 into @xmath226 which is still discrete and faithful .
the ( unique ) irreducible component @xmath251 in the @xmath226character variety containing the character @xmath252 of @xmath250 has ( complex ) dimension @xmath4 .
furthermore , if @xmath253 are the boundary tori of a compact core of @xmath191 , and if @xmath254 is a non trivial element in @xmath255 which is carried by @xmath211 , then @xmath256 and @xmath252 is an isolated point of the set @xmath257 the respective irreducible components containing the so called _ complete representations _
@xmath250 and @xmath249 are denoted by @xmath258 and @xmath259 respectively . in particular , @xmath260 and @xmath261
are called the respective _ dehn surgery components _ of the character varieties of @xmath191 , since the holonomy representations of hyperbolic manifolds or orbifolds obtained by performing high order dehn surgeries on @xmath191 are near @xmath262 ( see @xcite ) .
given a presentation of @xmath255 with a finite number , @xmath263 , of generators , @xmath264 , we introduce four affine coordinates ( representing matrix entries ) for each generator , which we denote by @xmath265 for @xmath266 and @xmath267 .
we view @xmath240 as a subvariety of @xmath268 defined by an ideal @xmath7 in @xmath269 $ ] .
there are elements @xmath270 $ ] for each @xmath271 such that @xmath272 for each @xmath273 .
we identify @xmath209 and @xmath210 with generators of @xmath274 .
thus , @xmath209 and @xmath210 are words in the generators for @xmath255 , and we define the following polynomial equations in the ring @xmath275 $ ] : @xmath276 for @xmath277 .
let @xmath278 be the variety in @xmath279 defined by @xmath7 together with the above equations .
for each @xmath280 , the equations ( [ ev : cood m] [ ev : cood ml ] ) have a solution since commuting elements of @xmath226 always have a common invariant subspace . the natural projection @xmath281 is therefore onto , and @xmath282 is a dominating map .
if @xmath283 , then there is an action of @xmath284 on the resulting points @xmath285 by inverting both entries of a tuple @xmath286 to @xmath287 .
the group @xmath284 acts transitively on the fibres of @xmath282 .
the map @xmath282 is therefore finite to one with degree @xmath288 .
the maximal degree is in particular achieved when the interior of @xmath191 admits a complete hyperbolic structure of finite volume , since theorem [ thurston thm ] implies that there are points in @xmath289 . the _ eigenvalue variety _
@xmath290 is the closure of the image of @xmath278 under projection onto the coordinates @xmath291 .
it is therefore defined by an ideal of the ring @xmath292 $ ] in @xmath293 .
note that this construction factors through a variety @xmath294 , which is the character variety with its coordinate ring appropriately extended , since we can choose coordinates of @xmath241 such that @xmath295 ( as a function on @xmath241 ) is an element of the coordinate ring of @xmath241 for each @xmath271 .
we let @xmath296 be the natural quotient map , which is equal to @xmath297 on the first @xmath298 coordinates , and equal to the identity on the remaining @xmath299 coordinates .
there also is a restriction map @xmath300 , which arises from the inclusion homomorphisms @xmath301 , and we therefore denote the map @xmath302 by @xmath303 .
denote the closure of the image of @xmath104 by @xmath304 .
there is the following commuting diagram of dominating maps : @xmath305 note that the maps @xmath306 and @xmath307 have the property that every point in the closure of the image has a preimage , and that the maps @xmath308 and @xmath309 are all finite to one of the same degree .
recall the construction by culler and shalen .
we start with a curve @xmath310 and an irreducible component @xmath311 in @xmath240 such that @xmath312 to obtain the tautological representation @xmath313 .
let @xmath314 be an irreducible component of @xmath278 with the property that @xmath315 .
then @xmath316 is a finitely generated extension of @xmath317 , and we may think of @xmath318 as a representation @xmath319 . if @xmath320 contains a curve @xmath321 ,
then @xmath316 is a finitely generated extension of @xmath322 , and hence to each ideal point of @xmath321 we can associate essential surfaces using @xmath318 . since the eigenvalue of at least one peripheral element blows up at an ideal point of @xmath321 , the associated surfaces necessarily have non empty boundary ( see @xcite ) . thus ,
if there is a closed essential surface associated to an ideal point of @xmath113 , then either @xmath320 is 0dimensional , or there is a neighbourhood @xmath323 of an ideal point @xmath54 of @xmath314 such that there are ( finite ) points in @xmath324 .
the later are called _ holes in the eigenvalue variety _ , examples of which are given in @xcite .
we denote the logarithmic limit set of @xmath290 by @xmath325 .
there are symmetries of the eigenvalue variety which give rise to symmetries of its logarithmic limit set . if @xmath326 , so is @xmath327 for any @xmath328 . thus , if @xmath329 , then @xmath330 for @xmath277 .
if we factor by these symmetries , we obtain a quotient of the logarithmic limit set in @xmath223 .
the quotient map extends to a map @xmath331 of spheres , which has degree @xmath332 .
let @xmath333 and let @xmath334 be the block diagonal matrix with @xmath4 copies of @xmath335 on its diagonal .
then @xmath334 is orthogonal , and its restriction to @xmath336 is a map of degree one .
[ boundary slopes lemma ] let @xmath191 be an orientable , irreducible , compact 3manifold with non empty boundary consisting of a disjoint union of @xmath4 tori .
if @xmath337 is a point with rational coordinate ratios , then there is an essential surface with boundary in @xmath191 whose projectivised boundary curve coordinate is @xmath338 .
note that the composite @xmath339 is a smooth map of degree @xmath332 . since the set of rational points is dense in the logarithmic limit set , the image @xmath340 is a closed subset of @xmath222 . for a manifold @xmath191 with boundary consisting of a single torus , lemma [ boundary slopes lemma ]
coincides with the result of @xcite that slopes of sides of the newton polygon of the @xmath0polynomial are strongly detected boundary slopes of @xmath191 . in this case
, we may normalise @xmath222 and @xmath341 such that their elements are pairs of coprime integers .
assume that there is a side of the newton polygon of slope @xmath342 where @xmath343 with outward pointing normal vector @xmath344 . then @xmath345 according to section [ spherical duals ] .
we now perform matrix multiplication : @xmath346 and arrive at a point @xmath347 \in \bc ( m)$ ] .
the analysis for the remaining cases is similar , and we have recovered the known relationship .
the coordinate ratios of @xmath54 are rational , and hence lemma [ curve finding lemma ] provides a curve @xmath113 in @xmath290 such that @xmath114 . from description ( [ log lim def2 ] ) of elements in the logarithmic limit set , we know that @xmath54 encodes the definition of a real valued valuation on the function field of @xmath113 : @xmath348 since the coordinate ratios of @xmath54 are rational , we may rescale the valuation by @xmath349 , such that for all @xmath328 , @xmath350 and @xmath351 are integers , and such that there is @xmath352 with @xmath353 .
thus , @xmath354 is a normalised , discrete , rank 1 valuation of @xmath355 .
we finish the proof with standard arguments . for details
see section 2.1 of @xcite , the proof of proposition 2.3 of @xcite or section 5.6 of @xcite .
the construction by culler and shalen yields an action of @xmath255 on a simplicial tree @xmath356 associated to a finitely generated extension @xmath357 of @xmath355 and to a valuation @xmath358 for some @xmath359 .
the translation length of @xmath271 with eigenvalue @xmath360 ( as an element of @xmath357 ) is @xmath361 ( see @xcite , section 3.9 ) .
a dual essential surface @xmath190 associated to the action can be chosen such that the geometric intersection number of a loop in @xmath191 representing @xmath362 with @xmath190 is equal to @xmath363 for all @xmath364 .
this implies that a boundary curve of @xmath190 on @xmath211 is homotopic to one of @xmath365 , where @xmath366 and @xmath367 and @xmath368 are coprime .
we need to determine the relative sign .
if @xmath214 represents an element of the subgroup @xmath369 , it is contained in the stabilizer of an edge , and hence has translation length equal to @xmath127 since the action is without inversions . the elements @xmath209 and
@xmath370 translate points in opposite directions along their axis .
since @xmath210 commutes with @xmath209 , they have a common axis .
thus , the curve @xmath214 can have zero translation length only if @xmath209 and @xmath210 translate in opposite directions .
this is the case if and only if @xmath371 and @xmath372 have opposite signs .
thus , the relative sign of @xmath218 and @xmath219 is opposite the relative sign of @xmath371 and @xmath372 .
we conclude that a boundary curve of @xmath190 on @xmath211 corresponds to the unoriented homotopy class of the element @xmath373 .
thus , the boundary curves of @xmath190 are encoded by : @xmath374 this completes the proof .
we now deduce some bounds on the dimension of the eigenvalue and the character variety .
[ ev : dim pro ] let @xmath191 be an orientable , irreducible , compact 3manifold with boundary consisting of @xmath4 tori
. then @xmath375
. moreover , if the interior of @xmath191 admits a complete hyperbolic structure of finite volume , then @xmath376 .
we have @xmath377 , where the first inequality is due to theorem [ ev : dim bc ] , the second to lemma [ boundary slopes lemma ] , and the equality to theorem [ bergman - bieri - groves ] .
it follows that @xmath378 . to prove the second statement , it is enough to show that under the assumptions , there is a subvariety of @xmath290 of dimension @xmath379 .
theorem [ thurston thm ] asserts that there is a component @xmath380 of @xmath241 of complex dimension equal to @xmath4 .
thurston shows in @xcite , section 5.8 , that the map @xmath381 defined by @xmath382 maps a small open neighbourhood of @xmath252 to an open neighbourhood of @xmath383 .
this is only possible if @xmath384 are algebraically independent over @xmath140 as elements of @xmath385 $ ] .
there is a component @xmath386 of @xmath294 with the same properties , since @xmath387 is a finite dominating map and the induced homomorphism @xmath388 \to \c [ x'_0]$ ] is injective .
let @xmath389 .
if the functions @xmath390 , @xmath277 , are algebraically dependent over @xmath140 as elements of @xmath391 $ ] , then they are also algebraically dependent over @xmath140 as elements of @xmath392 $ ] since we have a ring homomorphism @xmath391 \to \c [ x'_0]$ ] .
but the identities ( [ ev : cood m ] ) imply that then the functions @xmath393 are algebraically dependent over @xmath140 as elements of @xmath392 $ ] , contradicting the fact that @xmath394 is injective .
thus , the functions @xmath395 , @xmath277 , are algebraically independent over @xmath140 as elements of @xmath391 $ ] , and hence @xmath396 according to ( @xcite , 9.5 theorem 2 ) .
this completes the proof .
let @xmath191 be an orientable , irreducible , compact 3manifold with boundary consisting of @xmath4 tori . if @xmath191 contains no closed essential surface , then @xmath397 .
we have @xmath398 .
assume there is a component @xmath152 of @xmath294 such that @xmath399 .
then @xcite , theorem ( 3.13 ) , implies that there is a subvariety @xmath1 of dimension @xmath400 in @xmath152 which maps to a single point of @xmath290 .
associated to an ideal point of a curve in @xmath1 , we find a closed essential surface since the traces of all peripheral elements are constant on @xmath1 .
this contradicts our assumption on @xmath191 .
thus , for all components @xmath152 of @xmath241 , we have @xmath401 , and this proves the claim .
let @xmath191 be an orientable , irreducible , compact 3manifold with boundary consisting of @xmath4 tori , such that the interior of @xmath191 admits a complete hyperbolic structure of finite volume .
the proof of proposition [ ev : dim pro ] implies that the restriction of @xmath104 to the dehn surgery component , @xmath402 , is finite to one onto its image .
the degree of @xmath403 depends upon @xmath404 , so let us consider the corresponding map @xmath405 . in the case
where @xmath406 , dunfield has shown in @xcite that @xmath407 has degree 1 onto its image using thurston s hyperbolic dehn surgery theorem and a volume rigidity theorem ( gromov thurston
goldman ) . it would be interesting to generalise dunfield s proof to arbitrary @xmath4 , showing that @xmath407 is always a birational isomorphism onto its image .
analogous to @xmath409 , one can define a @xmath109eigenvalue variety @xmath410 , since the function @xmath411 defined by @xmath412 is regular ( i.e. polynomial ) for all @xmath413 ( see @xcite ) .
thus , we construct a variety @xmath414 using the relations @xmath415 for @xmath277 .
consider the representation of @xmath416 into @xmath109 generated by the images of @xmath417 in @xmath109 , the image of this representation is isomorphic to @xmath418 , but the image of any lift to @xmath226 is isomorphic to the quaternion group @xmath419 ( in quaternion group notation ) .
if @xmath420 restricts to such an irreducible abelian representation on a boundary torus @xmath211 , then the equations @xmath421 , @xmath422 , @xmath423 have no solution .
in particular , the projection @xmath424 may not be onto .
the closure of the image of @xmath414 onto the coordinates @xmath425 is denoted by @xmath426 and called the _
@xmath109-eigenvalue variety_. the relationship between ideal points of @xmath426 and strongly detected boundary curves is the same as for the @xmath226version , since the proof of lemma [ boundary slopes lemma ] applies without change : [ psl boundary slopes lemma ] let @xmath191 be an orientable , irreducible , compact 3manifold with non empty boundary consisting of a disjoint union of @xmath4 tori . if @xmath427 is a point with rational coordinate ratios , then there is an essential surface with boundary in @xmath191 whose projectivised boundary curve coordinate is @xmath338 .
in general , one needs a computer to calculate the @xmath0polynomial .
it is shown in @xcite that if @xmath428 is a nontrivial @xmath429torus knot , then @xmath430 is divisible by @xmath431 ( where we follow the convention that if @xmath191 is the complement of a knot @xmath428 in @xmath432 , then @xmath433 is a standard peripheral system ) .
the fundamental group of @xmath438 is presented by @xmath439 , and a standard peripheral system is given by @xmath440 and @xmath441 , where @xmath442 .
these facts can be found in @xcite on page 45 .
the factor @xmath443 arises from reducible representations , and all factors arising from components containing irreducible representations are to be determined .
the element @xmath444 is in the centre of @xmath445 , since it is identical to @xmath446 and hence commutes with both generators .
thus the image of @xmath444 is in the center of @xmath447 .
since two commuting elements have a common eigenvector , @xmath448 has a common eigenvector with @xmath449 and @xmath450 respectively .
if the representation is irreducible , these eigenvectors have to be distinct .
thus , after conjugation it may be assumed that the generators map to upper and lower triangular matrices respectively , and the central element is represented by a diagonal matrix .
direct matrix computations show that the commutativity with either of the generators gives @xmath451 .
assume that @xmath452 .
the relation @xmath453 , which is equivalent to @xmath454 , then implies the equation @xmath455 .
this is the curve obtained in @xcite by sending @xmath456 and @xmath16 to noncommuting elements of @xmath457 of order exactly @xmath458 and @xmath459 respectively . if the image of @xmath460 is trivial , then @xmath461 .
if @xmath462 or @xmath463 equals @xmath464 , this implies that the image of one of the generators is @xmath465 .
but this yields that @xmath447 is abelian , and therefore contradicts the irreducibility assumption .
this completes the proof of the first assertion .
now assume that neither of @xmath462 and @xmath463 is equal to two .
as above , curves of irreducible representations are obtained by sending @xmath456 and @xmath16 to noncommuting elements of @xmath457 of order exactly @xmath462 and @xmath463 respectively .
any of these curves yields a component of the eigenvalue variety defined by the equation @xmath466 , and this finishes the proof of the proposition .
in fact , with a more precise analysis one could count the number of 1dimensional curves in the character variety of a torus knot . similarly to the above , we can compute a generator for the defining ( principal ) ideal of the @xmath109eigenvalue variety of a torus knot .
we call this generator the @xmath467polynomial , and its variables have the unfortunate names @xmath468 and @xmath191 .
thus , @xmath469 is the @xmath467polynomial associated to the manifold @xmath191 .
it follows from the lemma in section 6 of @xcite that all @xmath109representations of a knot group lift to @xmath226 .
thus , all factors of the @xmath0polynomial arise from factors of the @xmath467polynomial and vice versa .
the eigenvalue of the meridian appears with even powers in the ( unfactorised ) @xmath0polynomial of @xmath428 , and the sign of the longitude s eigenvalue is uniquely determined since @xmath471 is null homologous .
thus , @xmath472 , but @xmath473 does not imply @xmath474 in general .
however , since @xmath475 , it follows that the variety defined by @xmath476 is the locus of @xmath477 .
expanding @xmath478 , substituting @xmath479 and @xmath480 and deleting repeated factors therefore gives the result as stated .
the contents of this paper forms part of my ph.d .
thesis , and i thank my supervisors craig hodgson and walter neumann for their constant support and inspiration .
this research was supported in part by an international postgraduate research scholarship by the commonwealth of australia department of education , science and training . | the @xmath0polynomial of a manifold whose boundary consists of a single torus is generalised to an _
eigenvalue variety _ of a manifold whose boundary consists of a finite number of tori , and the set of strongly detected boundary curves is determined by bergman s _ logarithmic limit set _ , which describes the exponential behaviour of the eigenvalue variety at infinity .
this enables one to read off the detected boundary curves of a multi
cusped manifold in a similar way to the 1cusped case , where the slopes are encoded in the newton polygon of the @xmath0polynomial . |
the properties of polymers in solution are determined by the balance between effective monomer - monomer attractions and excluded volume repulsions . upon increasing the relative strength of the attractions , for example by cooling ,
several regimes are encountered . as long as the thermal energy @xmath0 far exceeds the absolute value @xmath1 of the effective monomer - monomer attraction ,
good solvent conditions prevail , and individual polymer coils are swollen due to the dominance of excluded volume effects . the polymer radius of gyration @xmath2 scales like @xmath3 , where @xmath4 is the number of monomers or ( kuhn ) segments , and @xmath5 is the flory exponent@xcite .
when the temperature is lowered , the coils shrink due to the action of effective ( solvent - induced ) monomer - monomer attraction , until the theta regime is reached ; at the theta - temperature @xmath6 , monomer - monomer repulsion and attraction cancel , at least at the two - body level , and individual coils behave essentially like ideal ( random walk ) polymers , such that their radius of gyration scales like @xmath7 .
below @xmath6 individual coils collapse into dense globules with @xmath8 . moving away from the infinite dilution limit ,
interactions between different polymer coils come into play , and give rise to the lowest order correction to the van thoff limit of the osmotic equation of state , valid for non - interacting polymers ( see eq .
( [ eqhoff ] ) below ) .
the correction , due to pair interactions , is proportional to the square of the overall monomer concentration @xmath9 ; the coefficient is the second virial coefficient @xmath10 .
the boyle temperature @xmath11 is the temperature at which @xmath12 vanishes , i.e. @xmath13 .
since this condition is met when the monomer - monomer repulsion and attraction cancel , one expects that @xmath14 . finally , upon lowering the temperature at a finite polymer concentration , the polymer solution is found to separate into polymer - poor and polymer - rich phases below a critical temperature @xmath15 . within flory - huggins mean - field theory @xmath15
is found to coincide with @xmath11 within corrections of order @xmath16 .
thus , a polymer solution is characterized by three temperatures @xmath6 , @xmath11 , and @xmath15 , of which the first is a single polymer property , defined in the @xmath17 limit . for any
given @xmath4 , it is believed that @xmath18@xcite , but in the scaling limit @xmath17 , @xmath19 .
in fact , when simulating simple models of non - ideal polymers , probably the most accurate estimate for @xmath6 in the scaling limit is obtained by extrapolating results for @xmath20 to @xmath21@xcite . in this paper the operational definition of @xmath6 will be @xmath22 for sufficiently long chains
@xmath23 .
schematic phase - diagram of a polymer solution as a function of temperature @xmath24 and monomer concentration @xmath9 , for a finite length @xmath4 .
@xmath2 decreases with decreasing @xmath24 so that for good solvents , the monomer concentration at which the dilute regime crosses over to the semi - dilute regime increases .
as @xmath24 is lowered further , there is also a crossover to the `` theta regime '' , denoted schematically by the shaded region . in the limit @xmath25 , @xmath11 and @xmath15 both approach @xmath6 , so that the size of the `` theta regime '' decreases
. furthermore , the monomer concentration at the crossover between the dilute and semi - dilute regimes , as well as that of the critical point , both tend to zero for infinite @xmath4 .
, width=302 ] for a finite @xmath4 , there is a `` theta solvent regime '' of temperatures around @xmath6 , where the polymer behavior most resembles that predicted for theta polymers .
the longer the polymer , the sharper the transition between `` theta solvent '' and `` good solvent '' regimes@xcite . a typical phase - diagram for solutions of polymers of finite length is presented in fig .
[ fig : schematic ] . in the low concentration limit
, the physical properties of interacting polymers in solution closely resemble those of ideal polymers with the same @xmath2 . for good solvents , however , important deviations from ideal behavior rapidly set in for increasing concentration , even in the dilute regime , see e.g. ref . @xcite . on the other hand ,
the fact that @xmath26 for @xmath24 near @xmath6 suggests that in the theta regime the osmotic pressure @xmath27 will follow van thoff s law @xmath28 for a larger concentration range than is found for polymers in good solvent . nevertheless , as concentration is increased further , the effect of higher order virial coefficients will eventually kick in , leading to deviations from the simple van thoff behavior .
similar differences with ideal - polymer behavior should also be observable for interfacial properties such as the adsorption or surface tension near a hard non - adsorbing wall . while much theoretical and numerical work on simple models of interacting polymer solutions has been devoted to behavior for athermal conditions , or at @xmath29 , less
is known about how properties of polymer solutions vary with temperature and concentration in the intermediate regime @xmath30 . in the present paper we present mc data for the bulk and interfacial properties of dilute and semi - dilute polymer solutions for the same model as used in ref@xcite , over a range of temperatures between the athermal ( high temperature ) limit and @xmath6 .
the main objective is a systematic investigation of polymer solution properties over a wide range of concentrations , to determine how equilibrium properties like the osmotic pressure , the adsorption near a wall , the surface tension and the polymer induced depletion potential between two walls vary as thermal conditions change from the good solvent to the theta - solvent regime .
in particular we are examining the important problem of how the behavior of polymers in theta - solvent differs from that of ideal ( non - interacting ) polymers as their concentration increases .
the paper is organized as follows . the polymer lattice model and the computational methodology
are introduced in section ii .
results for the osmotic equation of state and semi - dilute regimes will be presented and discussed in section iii .
we shall next turn our attention to monomer density profiles near a non adsorbing wall as well as the polymer - wall surface surface tension as a function of temperature and concentration ( section iv ) .
the polymer - induced depletion potential between two walls will be discussed in section vi , and concluding remarks will be made in section vii .
a familiar coarse - grained representation of linear polymers in a solvent , which captures the essential physical features , is a self - avoiding walk ( saw ) lattice polymer , where each lattice site can be occupied by at most one monomer ( or polymer segment ) to account for excluded volume , and where pairs of non - sequential monomers of nearest - neighbor ( n.n . )
sites experience an attractive energy @xmath31 .
the parameter @xmath1 accounts for the difference between solvent - solvent , solvent - monomer , and monomer - monomer interactions , and includes both energetic and entropic components@xcite . as the dimensionless ratio @xmath32 increases from zero ( corresponding to the athermal solvent limit ) to larger values , the quality of the solvent decreases , i.e. effective monomer - monomer attractions become increasingly important , so that the polymer coils tend to shrink until the theta - regime is reached . upon further cooling at finite concentrations , phase separation sets in at some @xmath33 , see e.g. fig . [
fig : schematic ] . in this paper
we consider polymer chains made up of @xmath4 monomers in a simple cubic lattice of @xmath34 sites ( coordination number @xmath35 ) , with periodic boundary conditions . if @xmath36 polymers live on that lattice , the polymer concentration is @xmath37 , while the monomer concentration is @xmath38 .
@xmath39 , ( with @xmath40 in good solvent , and @xmath41 in theta solvent ) is the radius of gyration of a polymer coil , and the overlap concentration is defined as @xmath42 , where @xmath2 , which depends on @xmath43 , will conventionally be chosen to be the radius of gyration in the infinite dilution limit ( @xmath44 ) limit .
@xmath45 separates the dilute regime of the polymer solution ( @xmath46 ) from the semi - dilute regime @xmath47 . in the scaling limit @xmath48
, the properties of a polymer solution in the dilute and semi - dilute regimes depend only on @xmath49 and @xmath2 , and are independent of the monomer concentration @xmath9@xcite . in other words , simulations with different @xmath9 but
the same @xmath49 should give equivalent results when length is expressed in units of @xmath2 .
this requires that simulations be carried out with sufficiently long polymers so that the monomer concentration @xmath50 . under good solvent conditions ,
@xmath51 for polymers on a simple cubic lattice , so that @xmath52 whereas in theta - solvent , @xmath53@xcite , and so @xmath54 most of the subsequent mc simulations have been carried out for @xmath55 polymers , such that @xmath56 under good solvent conditions , while @xmath57 in theta - solvent . in order to ensure that @xmath58 in all simulations , the semi - dilute regime which could be explored was hence restricted to @xmath59 in good solvent , and @xmath60 in theta - solvent . for higher concentrations ,
the c - dependence of the results would no longer be negligible , and non - universal effects would become important , depending on the property under consideration .
for this reason , much smaller polymers , say @xmath61 , could not be used to study the semi - dilute regime .
the previous extensive mc study by grassberger and hegger@xcite determined @xmath11 as a function of @xmath4 for a model identical to the one used here .
they found that for @xmath55 , the boyle temperature @xmath11 gave @xmath62 , the value which we shall use throughout the discussion in this paper as a reasonable estimate for the theta - temperature .
we used 4 different types of moves to sample configuration space : the pivot algorithm attempts to rotate part of the polymer around a randomly chosen segment .
translations move the whole polymer through space , and reptations remove a segment at one end of the polymer and attempt to regrow it at the other end . since translation and pivot rapidly become less efficient for increasing polymer concentration , we also used configurational bias monte carlo ( cbmc ) moves@xcite . these are also very effective at lower temperatures , since the attraction between polymers reduces the accepted number of pivot and translation moves@xcite .
simulations were carried out in a box of size @xmath63 for polymers of length @xmath55 . for calculations of radii of gyration , and other bulk properties ,
periodic boundary conditions were used in all directions , whereas for surface properties calculations , hard walls were placed at @xmath64 and @xmath65 along the x - axis .
the osmotic pressures were calculated using the method of dickman@xcite . a external repulsive potential @xmath66 that acts between on monomers next to the wall is introduced , and varied such that the parameter @xmath67 .
@xmath68=0 prevents any particles from being adjacent to the wall , and the varying wall hardness allows the the osmotic pressure @xmath69 to be calculated from the integral @xmath70 where @xmath71 is the site occupation fraction of monomers right next to the wall .
we performed @xmath72 simulations at different @xmath68 values , chosen from a standard gauss - legendre distribution , to carry out the integral above . by using 2 walls of varying hardness ,
the statistics are enhanced over using a single wall .
adding repulsive walls increases the bulk density at the center of the simulation cell , an effect which becomes more pronounced for smaller boxes .
thus @xmath73 should not be taken as the bulk density .
instead , we have determined the bulk density by averaging over the region , near the center of the box , where the influence of the two depletion layers have completely decayed@xcite . as a first application of the mc code ,
we have computed the radius of gyration of an isolated polymer for lengths @xmath61 to @xmath74 , and extracted an exponent @xmath75 , shown in fig .
[ fig : lvsr ] . for higher temperatures ,
@xmath76 within error bars , as expected for the good solvent regime , but for @xmath77 the exponent @xmath78 appears to deviate slightly , and we find @xmath79 . in the limit @xmath17
, this scaling should be independent of temperature , as long as @xmath80 .
however , here we observe some rounding , most likely due to finite length crossover effects .
the precise temperature at which theta - type scaling behavior of @xmath2 sets in is dependent on length .
the longer the polymer , the smaller the range of temperatures over which transition from the good - solvent to the theta regime occurs@xcite .
note also that for a given @xmath4 , the @xmath2 for e.g. @xmath81 , is smaller than @xmath2 for @xmath82 .
nevertheless , the scaling with length , @xmath83 , is still the same for those two temperatures , i.e. the polymers are still in the good solvent regime .
thus the effect of lowering the temperature is simply to renormalize the effective step - length . for @xmath84 , with @xmath11 for each length taken from ref .
@xcite , we find @xmath85 , as expected for the theta regime@xcite . as the temperature
is lowered further , @xmath78 continues to decrease , and for low enough @xmath24 the polymer should collapse to a compact globule state where @xmath86 .
the beginning of this trend is evident in the plot for @xmath87 .
the larger error bars reflect the more pronounced finite @xmath4 effects expected at this temperature . for short polymers
the transition from extended to collapsed states occurs over a broad range of temperatures , whereas for longer polymers the collapse transition is sharper@xcite .
effective exponents @xmath78 , extracted from simulations with different lengths @xmath4 , as a function of @xmath43 . .
note that the value presented at @xmath88 , corresponding to @xmath11 for @xmath55 , was calculated at slightly different temperatures , corresponding to @xmath11 , for each polymer length.,width=302 ] @xmath2 shrinks ( for fixed @xmath4 ) as the temperature is lowered , because the excluded volume interactions are partially compensated by attractive interactions between monomers .
similarly in the good solvent regime , @xmath2 decreases with density because the excluded volume interactions are screened by overlapping polymer coils .
this decrease follows a scaling law @xmath89@xcite once the semi - dilute regime is well developed@xcite . at the theta - temperature ,
on the other hand , chain statistics are nearly ideal on all length scales and concentrations@xcite , so that @xmath2 should be independent of concentration . for even lower temperatures , @xmath90 ,
the screening of attractive interactions now implies that @xmath2 should increase with concentration .
these trends are indeed observed in our mc simulations , as illustrated in fig .
[ fig : neatrg ] for @xmath55 polymers .
@xmath2 , calculated for @xmath55 polymers , decreases with concentration for polymers in good solvent , is virtually independent of density for @xmath84 , and increases with concentration when @xmath91 .
for each temperature simulations were performed up to a maximum monomer concentration @xmath92 , so that the total @xmath49 range accessible decreases with temperature .
, width=302 ]
the thermodynamic property which is most readily extracted from the simulations is the internal energy of polymer solutions , which is simply @xmath93 , where @xmath94 is the average number of nearest - neighbor `` contacts '' between non - connected monomers .
the energy per polymer is @xmath95 , where @xmath96 is the average number of non - connected nearest neighbors around a monomer . within a simple mean - field picture , @xmath96
is expected to increase linearly with the polymer concentration . such linear behavior
is indeed observed in fig .
[ fig : energies ] for theta conditions . at higher temperatures
, there is a significant downward curvature on @xmath97 , while below @xmath11 , but above @xmath15 , @xmath97 curves upward .
mc data for the reduced internal energy @xmath98 for @xmath55 chains , plotted as a function of polymer concentration @xmath49 for 5 different temperatures .
note how the curvature changes with solvent quality.,width=302 ] linear graph of the eos @xmath99 as a function of concentration for different temperatures .
( inset ) a blowup of the dilute regime highlights the differences between near theta solvents and polymers in good solvent .
the former show behavior that mimics that of ideal polymers.,width=302 ] the eos as a function of concentration for different temperatures , plotted on a double logarithmic scale . in the semi - dilute regime ,
the eos is consistent with scaling theory , which predicts @xmath100 for good solvents and @xmath101 for theta solvents .
, width=302 ] the osmotic equation of state ( eos ) @xmath102 was calculated using dickman s method@xcite described in the previous section .
the mc results for @xmath103 as a function of @xmath49 , for various temperatures , are plotted in figures [ fig : neatopress ] and [ fig : neatlogplot ] on linear and logarithmic scales .
as expected , z increases with @xmath49 , and decreases with temperature@xcite .
since the theta - solvent regime is defined by @xmath104 , the eos is expected to be very flat at low polymer concentrations since @xmath11 is defined as the temperature at which the second virial coefficient @xmath105 vanishes in the virial expansion@xcite : @xmath106 this is indeed confirmed by the mc results in the dilute regime , as illustrated in the inset of fig .
[ fig : neatlogplot ] . up to @xmath107 ,
@xmath103 hardly rises with @xmath108 , i.e. the system behaves like a solution of ideal polymers . at higher concentrations
, however , interactions between monomer triplets come into play , so that , according to eq . ( [ eos ] )
, @xmath103 is expected to increase like @xmath109 , in agreement with a standard scaling argument@xcite .
the same scaling argument predicts that the eos of polymers in good solvent scales as @xmath110 ; the exponent @xmath111 indeed agrees with the slopes of the double logarithmic plots of @xmath112 shown in fig .
[ fig : neatlogplot ] , for @xmath113 and @xmath114 .
the slope of the theta - temperature eos is clearly significantly larger , and compatible with the expected exponent of @xmath115 .
however , for the @xmath55 chains used in the present simulations , the accessible semi - dilute regime of concentrations is too small to allow a fully quantitative estimate ( c.f . the discussion in section ii ) .
since the eos of theta - polymers scales with a larger exponent than that of polymers in good solvent , there will eventually be a concentration @xmath49 where the eos of the theta - polymers becomes larger than that of polymers in good solvent .
this may , at first sight , seem counter - intuitive , but it should be remembered that for a given size @xmath2 , theta - polymers are more compact . by extrapolating our simulation data to larger @xmath49 , we expect the cross - over between theta - polymers and @xmath116 ( saw ) polymers to occur at @xmath117 which is , in practice , very hard to achieve both in simulations and experiments@xcite . on the other hand ,
if one starts from a solution under good solvent conditions , i.e. `` high '' temperature , at a given @xmath108 , and lowers the temperature , then the eos will decrease monotonically , because @xmath2 , and with it the ratio @xmath49 , decreases with temperature .
for entropic reasons , polymer solutions will be highly depleted in the vicinity of a non - adsorbing hard wall : since fewer conformations are available to a polymer coil the wall will effectively repel the polymers so that their concentration will drop from the bulk value for distances @xmath118 , to a much lower value at contact@xcite .
the case of polymers in good solvent ( in the saw limit ) has been investigated in much detail in ref .
@xcite , where important deviations from ideal polymer behavior were found even in the dilute regime . here
we continue these investigations , but for polymers in solvents of varying quality .
fig [ fig : neatdprofeps2 ] compares the present mc results for the reduced monomer density profiles @xmath119 as functions of the distance @xmath120 from the planar surface in the good solvent @xmath121 ) and theta - solvent ( @xmath122 ) limits for @xmath55 chains , and several bulk concentrations .
the differences between the two regimes are very significant . in the dilute regime
the density profiles in the theta solvent hardly change , and are remarkably close to that of ideal polymers , while the profiles in good solvent already differ significantly from the latter , and are `` pushed '' closer toward the hard wall . in the semi - dilute regime , on the other hand , the density profiles of polymers in theta - solvent differ considerably from that of ideal polymers ( which is independent of concentration ) and instead more closely resemble those of polymers in good solvent , although the depletion `` hole '' is systematically wider for the former compared to that of the latter .
this is illustrated in fig .
[ fig : neatdprof ] , where the profiles at three different temperatures are compared at low concentration , and in the semi - dilute regime .
the reduced adsorption of polymers near a hard wall is defined as@xcite : @xmath123 where @xmath124 is the surface excess grand potential per unit area @xmath125 , @xmath126 is the chemical potential , and @xmath127 . because of the depletion hole , @xmath128 is negative , and has the dimension of length .
one could replace @xmath129 by the distribution of the center of mass ( cm ) of the polymers @xmath130 .
although the reduced profiles @xmath131 would look different from @xmath132 ( see e.g. fig.1 of ref .
@xcite ) , the reduced adsorptions would be identical to those calculated from the monomer density profiles , due to the conservation of the number of monomers . for ideal polymers
, the reduced adsorption can be exactly calculated to be@xcite : @xmath133 which is independent of concentration . for polymers in good
solvent , a renormalization group ( rg)@xcite calculation predicts that @xmath134 for the adsorption in the low - density limit@xcite , which is close to , but slightly less negative than the ideal result ( [ eq2.3b ] ) . as the polymer concentration increases , interactions between monomers of non - ideal polymers cause the depletion layer to shrink , as illustrated by the density profiles shown in figs .
[ fig : neatdprofeps2 ] and [ fig : neatdprof ] .
thus , in contrast to the case of ideal polymers , where @xmath119 , and therefore @xmath135 is independent of concentration , for interacting polymers the reduced adsorption @xmath135 should become less negative for increasing polymer concentration .
this is indeed what is observed in fig .
[ fig : neatadsorb ] , where the variation of the reduced adsorption with concentration is shown for several temperatures . in the dilute regime @xmath136
, there is , once again , a qualitative difference between polymers in theta or good solvent , as highlighted in the inset of fig .
[ fig : neatadsorb ] . while the curvature of the @xmath128 curves with concentration is negative in good solvent conditions , it is clearly positive under theta - conditions , and changes to negative in the semi - dilute regime . up to @xmath107 ,
the results for @xmath135 remain close to ( but above ) the ideal polymer result ( [ eq2.3b ] ) .
note , however , that the relative deviation from ideal polymers is more pronounced than for the osmotic eos ( compare the insets to figs . [
fig : neatopress ] and [ fig : neatadsorb ] ) , hinting at a stronger effect of three - particle interactions on the adsorption of theta - polymers .
monomer density profiles @xmath119 for polymers near a hard non - adsorbing wall .
the upper panel shows the density profiles for polymers in good solvent , while the lower panel is for polymers at @xmath84 . at these near theta conditions , the density profiles in the dilute regime are nearly indistinguishable from those of ideal polymers , but
important deviations emerge in the semi - dilute regime .
, width=302 ] monomer density profiles as a function of temperature for two densities .
the upper panel is for @xmath137 , and the lower panel is for @xmath138 , well into the semi - dilute regime .
the density profiles at lower temperatures are more extended when plotted in terms of @xmath139 . , width=302 ] reduced adsorption @xmath140 as a function of concentration for several temperatures .
( inset ) blowup of the dilute regime .
note the different curvature at low densities .
the values at @xmath141 are take from theory to be @xmath142@xcite for the first three temperatures , and @xmath143 for @xmath84 .
, width=302 ] in the semi - dilute regime we expect the adsorption to be proportional to the correlation length @xmath144@xcite , since this is the only relevant length - scale .
note that the numerical prefactors for different semi - dilute correlation lengths vary with the property one is attempting to describe , see e.g. @xcite for a very useful discussion of this matter . in our previous paper@xcite we showed that for polymers in good solvent , the reduced adsorption @xmath145 is consistent with this scaling . here
we present more simulations in the semi - dilute regime to confirm this behavior , finding that @xmath146 provides a good fit .
scaling theory considerations also predict that @xmath147 for theta solvents@xcite .
however , this behavior can not be reliably extracted from the current simulations of @xmath135 , which are not performed for a large enough range of @xmath49 .
furthermore , we expect that for the higher values of @xmath49 sampled here , finite @xmath9 effects may already be coming into play .
the wall - polymer surface tension , i.e. the free - energy cost of introducing a non - adsorbing hard wall and its associated depletion layer , can be calculated from the adsorption and eos by use of the gibbs adsorption equation@xcite : @xmath148 by performing one integration by parts w.r.t .
density , eq .
( [ eq2.4 ] ) can also be expressed as : @xmath149 the first term in this equation has an appealing physical interpretation as the free energy cost per unit area of creating a slab cavity of width @xmath128 . for ideal polymers , where @xmath128 is independent of concentration
, this term is the only one that contributes , and so @xmath150 .
since the eos and @xmath128 for interacting polymers in various quality solvents were calculated in the previous sections , we can now , using eq .
( [ eq2.4b ] ) , determine the surface tension for different solvent qualities .
results are shown in fig .
[ fig : neatst3 ] . for good solvent conditions , i.e. @xmath151 , they are , to within simulation errors , in near quantitative agreement with the rg calculations of maasen , eisenriegler , and bringer@xcite .
in fact , the larger number of simulation results in the semi - dilute regime , used in the present work , lead to even better agreement than that shown earlier in ref . @xcite .
we are not aware of theories of comparable quality for the surface tension of theta polymers .
as discussed earlier for the eos and the adsorption , we again find qualitatively different behavior for the @xmath84 solution in the dilute regime , as highlighted in the inset of fig .
[ fig : neatst3 ] .
the surface tension for such theta like solvents resembles that of ideal polymers much more closely than that of polymers in good solvent .
as expected from the related behavior of the adsorption , the relative deviation from ideal polymer behavior in the dilute regime is somewhat larger than what was found for the eos .
note that the curves @xmath152 cross at very low @xmath49 .
this follows from the fact that the low concentration limit of the surface tension is given by @xmath153 , and @xmath154 is smaller for interacting polymers than for ideal or theta polymers .
( see fig .
[ fig : neatadsorb ] ) . at slightly higher densities ,
the curves cross because @xmath155 increases more rapidly with concentration for good solvents than for theta solvents . in the semi - dilute regime , where @xmath156 and @xmath157 , eq .
( [ eq2.4 ] ) simplifies@xcite to @xmath158 this implies @xmath159 semi - dilute scaling for polymers in good solvent , and @xmath160 behavior for theta solvents , which is consistent with the results in fig .
[ fig : neatst3 ] .
furthermore , this scaling suggests that for large enough @xmath49 , the surface tension curve for theta polymers will eventually cross that of polymers in good solvent . indeed , fig .
[ fig : neatst3 ] already appears to suggest this behavior , although some care must be taken in extracting the @xmath49 at which crossover occurs , due to the the possibility of finite @xmath9 effects for theta polymers@xcite . with these caveats in mind ,
our results suggest that the surface tension for theta polymers would be greater than that of interacting polymers when @xmath161 , which is considerably lower than the predicted crossing for the eos , and within reach of experiment and simulations . on the other hand , just as in the case of the eos , if one were to lower the temperature of an experimental polymer solution at a given @xmath108 , this would lead to a lower @xmath49 , so that a monotonic behavior of @xmath155 with temperature is expected when following this route .
wall - polymer surface tension divided by the ideal surface tension for various solvent qualities , plotted as a function of concentration @xmath49 .
( inset ) a blowup of the dilute regime highlights the differences between @xmath162 for good and theta solvents.,width=302 ]
the depletion potential between two walls or plates induced by a polymer solution is defined as the difference in free energy between the cases where the plates are at a distance @xmath120 , or infinitely far apart .
the polymer solution between the plates is taken to be in equilibrium with a much larger reservoir of pure polymer solution at the same chemical potential . at infinite distance ,
the free energy cost of having two plates is simply twice the cost of making a depletion layer , while at contact , these two depletion layers are destroyed .
thus , when @xmath163 , the free energy per unit area of plate is given by @xmath164 for small @xmath120 virtually no polymer is expected to penetrate between the plates , which are hence pressed together by the osmotic pressure @xmath165 , leading to a linear increase of @xmath166 with @xmath120 . as discussed in ref @xcite ,
the simplest approximation for the depletion potential is to continue this linear form for larger @xmath120 : @xmath167 where the range is given by @xmath168 we note that this approximation is similar to that adopted by joanny , leibler , and de gennes who , in their pioneering paper@xcite , approximated the force between two plates as constant for @xmath169 and zero for @xmath170 .
this also results in a linear depletion potential . in ref @xcite
, we showed that this simple theory is virtually quantitative for polymers in good solvent@xcite .
furthermore , it is well known to be quite accurate for ideal polymers as well@xcite .
although one could easily generalize the theory to include the small curvature seen for ideal polymers close to @xmath171@xcite , we ignore these small corrections in the interests of simplicity . since this theory is accurate for polymers in good solvent as well as for ideal polymers , we postulate that the same simple assumptions are valid for other solvent qualities , and use eq .
( [ eq2.6 ] ) to calculate the depletion potentials .
the well - depth @xmath172 follows from eq .
( [ eq3.1 ] ) , and is shown in fig . [
fig : neatw0 ] .
the behavior mirrors that of fig .
[ fig : neatst3 ] of course . from the inset
it is clear that theta polymers most closely resemble ideal polymers in the dilute regime , while important deviations are found in the semi - dilute regime . the depletion potential at contact is given by @xmath173 .
( inset ) a blowup of the dilute regime shows that theta polymers most closely resemble the ideal polymers .
, width=302 ] in fig .
[ fig : neatdep2 ] , we plot the range @xmath174 for a number of different solvent qualities . in the dilute regime , highlighted in the inset , the range for the near theta solvent is fairly close to that of ideal polymers , whereas for polymers in good solvent the range decreases markedly in the dilute regime . in the semi - dilute regime
both good and theta solvent regimes show important deviations from ideal polymer behavior . from scaling theory
we expect @xmath175 and @xmath176 for the polymers in good and theta solvents respectively .
range of the depletion potential @xmath174 , as given by eq .
( [ eq2.6a ] ) . in the dilute regime ,
highlighted in the inset , theta polymers are most similar to ideal polymers , whereas polymers in good solvent already show fairly strong deviations : at @xmath177 the range has dropped by almost a factor 2 compared to ideal polymers . , width=302 ]
the results for @xmath172 and @xmath174 can be combined with eq .
( [ eq2.6 ] ) to determine the full depletion potential between two plates .
an example , for @xmath178 , is shown in fig .
[ fig : neatdeprho1 ] .
the depletion potential for theta polymers most closely resembles that of ideal polymers . for lower @xmath49
we expect an even closer correspondence .
depletion potentials , calculated via eq .
( [ eq2.6 ] ) , for @xmath179 and @xmath180 .
this simple theory compares well to direct simulations , taken from @xcite , for good solvents . for theta polymers the depletion potential resembles that of ideal polymers : at @xmath181 they are virtually inistinguishable ; at @xmath180 there are slight differences . , width=302 ] in the dilute regime , figs . [
fig : neatw0][fig : neatdeprho1 ] suggest that the depletion potential between two plates , induced by theta polymers , resembles that of ideal polymers .
we have recently shown how to construct depletion potentials between two spheres from that between two plates@xcite , and found quantitative agreement with direct simulations of @xmath55 saw polymers between spheres , and also good results for ideal polymers .
this suggests that the same procedure should work well for the depletion potential between two spheres induced by theta polymers .
furthermore , since the depletion potentials are the dominant determinant of depletion induced phase separation for @xmath182@xcite , we would expect the phase - behavior of theta polymers to closely follow that of ideal polymers , at least for these size ratios .
if there are slight deviations , then fig .
[ fig : neatdeprho1 ] suggests that the binodals would shift toward those of polymers in good solvent .
recent experiments@xcite of colloids mixed with polymers in a near theta solvent indeed show behavior that more closely resembles that of ideal polymers than that of polymers in good solvent .
we have performed extensive monte carlo simulations to investigate the bulk and interfacial properties of polymers in good solvent , and polymers in the theta regime . for infinite length @xmath4 , there would be a sharp transition between the behavior of polymers in good solvent , and those in a theta solvent .
the simulations were carried out for @xmath55 , so that continuous cross - over effects are to be expected@xcite .
working out the detailed crossover behavior would require many more simulations at different @xmath4 .
however , in practice , most experimentally investigated polymer solutions do not reach the scaling regime either . moreover
, we do observe significant qualitative differences between theta polymers , and polymers with weaker monomer - monomer attraction .
in contrast to polymers in good solvent , solutions of polymers in theta solvent are quite well described by ideal polymer theories throughout the entire dilute regime .
this works best for bulk properties like the eos , and slightly less well for interfacial properties such as the density profiles , adsorptions , and surface tension .
however , as predicted by theory@xcite , in the semi - dilute regime theta polymers also begin to to exhibit important deviations from ideal - polymers .
our simulations show behavior which is consistent with that predicted by scaling theories , but the concentration range we were able to investigate is not wide enough to unambiguously confirm the expected scaling behavior .
we also suggest that , for large enough concentrations , the values of various properties , including the eos , the adsorption , and the surface tension , will be larger for theta polymers than for polymers in good solvent at the same reduced concentration @xmath49 .
the close agreement between theories for ideal polymers , and our simulations of theta polymers in the dilute regime , suggests that the effective depletion pair potentials , and associated phase behavior , should also resemble those of ideal polymers , at least in the so - called `` colloid limit '' , where @xmath182 .
we thank p.g .
bolhuis and v. krakoviack for use of their computer codes , and for their help , and andrea pelissetto for valuable discussions concerning the scaling limit .
cia thanks the epsrc for a quota studentship , and aal thanks the royal society for their financial support . 99 p.g de gennes , _ scaling concepts in polymer physics _ , ( cornell university press , ithaca , 1979 )
. m. rubinstein and r. colby , _ polymer physics _ , ( oxford university press , oxford , 2003 ) .
h. frauenkron and p. grassberger , j. chem . phys . *
107 * , 9599 ( 1997 ) .
q. yan and j.j .
de pablo , j. chem .
phys . * 113 * , 1276 ( 2000 ) .
p. grassberger and r. hegger , j. chem . phys . *
102 * , 6881 ( 1995 ) .
louis , p.g .
bolhuis , e.j .
meijer and j.p .
hansen , j. chem .
phys . * 116 * , 10547 ( 2002 ) .
d. frenkel and b. smit , _ understanding molecular simulations _
( academic press , 1995 ) .
v. krackoviak , j .-
hansen , and a.a .
louis , phys .
e , * 67 * , 041801 ( 2003 ) .
r. dickman , j. chem . phys . * 87 * , 2246 ( 1987 ) . a recent more detailed analysis of the dickman method ( m.r
. stukan , v.a ivanov , m. muller , w. paul and k. binder , j. chem .
phys . * 117 * , 9934 ( 2002 ) ) , pointed out some significant sources of finite size effects .
the most important one comes from the increase in the effective density due to the depletion layers that develop upon the introduction of two repulsive walls .
_ advocate doing simulations in a grand - canonical ensemble .
however , since our boxes are larger than those used by the authors above , so that the effect is smaller , our approximation of the density by the average near the center of the box , far from the depletion layers , should be adequate .
et al . _ also noted that in the integral of eq .
( [ eq : dickman ] ) , the value of @xmath183 did not drop to zero as @xmath184 .
again , this effect is not observed in the present simulations , carried out at much lower monomer densities .
l. schfer _ excluded volume effects in polymer solutions _ , ( springer verlag , berlin , 1999 ) ) . because @xmath6 is a tricritical point@xcite , there should be logarithmic corrections to the scaling behavior .
renormalization group ( rg ) theory predicts@xcite @xmath185 but these corrections are of the order of our simulation error bars , and so could not be resolved .
bolhuis , a.a .
louis , j .-
hansen , and e.j .
meijer , j. chem . phys .
* 114 * , 4296 ( 2001 ) .
note that for @xmath186 , we would expect the eos to fall onto the same master curve for all temperatures @xmath187 in the good solvent regime .
the fact that they do not in figs .
[ fig : neatopress ] and [ fig : neatlogplot ] is most likely due to finite @xmath4 crossover effects , which are also evident in some of the other physical properties studied in the current paper .
the @xmath49 value at the purported crossing of the two equations of state is quite sensitive to ( small ) finite @xmath4 effects .
our estimate is in part based on preliminary @xmath188 results ; nevertheless it may still be subject to fairly large uncertainty .
rowlinson and b. widom , _ molecular theory of capilarity _ , ( oxford university press , oxford 1989 ) .
s. asakura and f. oosawa , j. chem . phys .
* 22 * , 1255 ( 1954 ) .
a. hanke , e. eisenriegler and s. dietrich , phys .
e. * 59 * , 6853 ( 1999 ) .
huang and t.a .
witten , macromolecules * 35 * , 10225 ( 2002 )
. y. mao , p. bladon , h.n.w .
lekkerkerker and m.e .
cates , mol .
phys . * 92 * , 151 ( 1997 ) r. maasen , e. eisenriegler , and a. bringer , j. chem . phys . * 115 * , 5292 ( 2001 ) . a.a .
louis , p.g .
bolhuis , e.j .
meijer and j .- p .
hansen , j. chem .
phys . * 117 * , 1893 ( 2002 ) .
j. f. joanny , l. leibler , and p.g . de gennes , j. polym .
* 17 * , 1073 ( 1979 ) , see also g.j .
fleer , j.m.h.m .
scheutjens , and b. vincent , acs symp .
* 240 * , 245 ( 1984 ) . for a related simple theory for the depletion potential , see e.g. r. tuinier , d.g.a.l .
aarts , h.h .
wensink , and h.n.w .
lekkerkerker , phys .
phys . * 5 * , 3707 ( 2003 ) .
meijer and d. frenkel , j. chem . phys .
* 100 * , 6873 ( 1994 ) .
b. rotenberg , j. dzubiella , j .-
hansen , and a.a .
louis , cond - mat/0305533 .
shah , y.l .
chen , k.s .
schweizer and c.f .
zukoski , j. chem . phys . * 118 * , 3350 ( 2003 ) . | the effect of solvent quality on dilute and semi - dilute regimes of polymers in solution is studied by means of monte carlo simulations . the equation of state , adsorptions near a hard wall , wall - polymer surface tension and effective depletion potentials are all calculated as a function of concentration and solvent quality .
we find important differences between polymers in good and theta solvents . in the dilute regime
, the physical properties for polymers in a theta solvent closely resemble those of ideal polymers . in the semi - dilute regime , however , significant differences are found . |
the origin of the morphology of galaxies is one of the mysteries of galaxy formation .
recently , there has been a great deal of effort to understand the origin of morphology by inspecting dependences of galaxy properties on environment .
this line of study dates back to the 1930 s ( e.g. hubble & humason 1931 ) when it was realized that clusters were dominated by ellipticals and lenticulars and that environmental factors played an important role in determining the morphology of galaxies .
oemler ( 1974 ) found the morphology - radius relation ; the late type galaxy fraction decreases with radius within a cluster .
this relation was confirmed by dressler ( 1980 ) who argued that the fraction of morphological types is a function of local galaxy density .
postman & geller ( 1984 ) extended this morphology - density relation down to the group environment .
gunn & gott ( 1972 ) argued that s0 galaxies in great relaxed clusters like coma were the result of spirals being stripped of gas by ram pressure stripping due to hot intra - cluster gas .
gott & thuan ( 1976 ) argued that elliptical galaxies were produced by larger initial density fluctuations which would have higher density at turn - around and where star formation would be completed before collapse .
such larger initial density fluctuations would be more likely in a region that would later turn into a high density environment .
environmental dependences of galaxy properties other than morphology have also been studied extensively .
recently , as large galaxy redshift surveys like the two degree field galaxy redshift survey ( 2dfgrs ; colless et al .
2001 ) and the sloan digital sky survey ( sdss ) have been completed , extensive studies have been made to accurately measure the environmental effects on various physical properties of galaxies ( goto et al .
2003 ; balogh et al .
2004a , b ; rojas et al . 2004 , 2005 ; hoyle et al . 2005 ; blanton et al . 2005a ; croton et al . 2005 ; tanaka et al . 2005 ; weinmann et al . 2006
; boselli & gavazzi 2006 ; among many others ) . works done by others on similar topics
will be discussed in each section below when the dependence of each of physical properties of galaxies on environment is studied .
the observed environment - galaxy property relations provide only indirect clues to the origin of morphology , but they certainly are important constraints on models for the physics of galaxy formation . in a companion paper
we study the relations among various physical properties of galaxies in the sdss spectroscopic sample ( choi et al .
2006 , hereafter paper i ) . in this paper
we extend our work to the environmental dependence of the physical parameters .
we accurately measure the morphology - density - luminosity relation from a series of volume - limited samples drawn from the sdss .
we also use morphological subsets of early and late type galaxies to measure the dependences of various physical properties of galaxies on environment once the morphology and luminosity are fixed . by studying the environmental dependence at different smoothing scales , we address the important question of whether galaxy morphology depends primarily on the large - scale environment in which the galaxy initially formed or on a smaller scale environment that may reflect the influence of later evolutionary effects such as galaxy - galaxy interactions .
we use a large - scale structure sample , dr4plus , of the sdss ( york et al .
2000 ; blanton et al .
2003a ; fukugita et al .
1996 ; gunn et al .
1998 , 2006 ; hogg et al . 2001 ; ivezi et al . 2004
; lupton et al .
2001 ; pier et al .
2003 ; smith et al . 2002
; stoughton et al .
2002 ; tucker et al .
2006 ) from the new york university value - added galaxy catalog ( nyu - vagc ; blanton et al .
this sample is a subset of the recent sdss data release 5 ( adelman - mccarthy et al .
our major sample of galaxies used here is a subset of the large - scale structure dr4plus ( lss - dr4plus ) sample referred to as `` void0 , '' which includes the main galaxies ( strauss et al .
2002 ) with apparent magnitudes in the range @xmath7 for the approximately 6% of targeted galaxies that lack a measured redshift because of fiber collisions , we assign the redshift of the nearest neighbor .
completeness of the sdss is poor for bright galaxies with @xmath8 because of spectroscopic selection criteria ( which exclude objects with very large flux within the fiber aperture ) and difficulties of automatically measuring photometric properties of very extended sources . due to the magnitude limits of the void0 sample ,
the full range of magnitude of the sample is only @xmath9 at a given redshift , and is even smaller than that in the case of a volume - limited sample with a finite range in redshift .
as described in detail in section 2.1 of paper i , we add the missing bright galaxies and thereby extend the magnitude range by using various existing redshift catalogs which we match to the sdss data . in total , 5195 bright galaxies are added to the void0 samples , which means our whole sample includes 317533 galaxies within our angular sample boundaries ( see fig .
volume - limited samples derived from the resulting catalog have nearly constant comoving number density of galaxies in the radial direction for redshifts @xmath10 .
we treat our final sample as effectively having no bright limit at redshifts greater than @xmath11 .
more details about this sample can be found in paper i. to study the effects of environment on galaxy properties it is advantageous for the observational sample to have the lowest possible surface - to - volume ratio , to minimize the boundary effects ( see park et al . [
2005 ] regarding similar effects in topology analysis ) .
for this reason we trim the dr4plus sample as shown in figure 1 , in which gray lines delineate our sample boundaries in the survey coordinate plane [ ( @xmath12 ) ; stoughton et al .
there are also three stripes in the southern galactic cap observed by sdss that we do not use because of their narrow angular extent .
these cuts leave approximately @xmath13 square deg@xmath14 in the survey region . within our sample boundaries
, we account for angular variation of the survey completeness by using the angular selection function defined in terms of spherical polygons ( hamilton & tegmark 2004 ) , which takes into account the incompleteness due to mechanical spectrograph constraints , bad spectra , or bright foreground stars .
we make two arrays of square pixels of size @xmath15 in the ( @xmath16 ) sky coordinates covering our analysis region , and store the angular selection function calculated using the mangle routine ( hamilton & tegmark 2004 ) . at the location of each pixel ,
the routine calculates the survey completeness in a spherical polygon formed by the adaptive tiling algorithm ( blanton et al .
2003a ) used for the sdss spectroscopy . the resulting useful area
( with nonzero selection function ) within the analysis regions is 1.362 sr . in our study
we use only volume - limited samples of galaxies , defined by absolute magnitude and redshift limits as shown in figure 3 of paper i. in addition to the volume - limited samples , labeled d1 to d5 , which are used to study the environmental effects , we also use the @xmath17 galaxy ( ls ) sample .
the ls sample is designed to include ` typical ' @xmath17 galaxies ( @xmath18 ) , and covers the full range of comoving distance , from @xmath19 to 314.0 @xmath6mpc , in which our analysis is made .
all measurements of local density in this paper use the ls sample to define the local environment ( see section 3 ) .
the definitions for all samples are summarized in table 1 .
we analyze our samples in the survey coordinate frame , and the cartesian coordinates are chosen to be @xmath20 the @xmath21-axis is at the center of the survey coordinate and radially outward , the @xmath22-axis is toward the @xmath23 direction , and the @xmath24-axis is at the @xmath25 direction .
lccccc d1 & @xmath26 & @xmath27 & @xmath28 & 20288 ( 6256 ) & 3.41 + d2 & @xmath29 & @xmath30 & @xmath31 & 32543 ( 11341 ) & 3.78 + d3 & @xmath32 & @xmath33 & @xmath34 & 49571 ( 19270 ) & 4.18 + d4 & @xmath35 & @xmath36 & @xmath37 & 74688 ( 33039 ) & 4.58 + d5 & @xmath38 & @xmath39 & @xmath40 & 80479 ( 39333 ) & 5.56 + ls & @xmath41 & @xmath39 & @xmath40 & 66176 ( 30715 ) & 5.94 + the physical parameters considered in this study are absolute petrosian magnitude in the @xmath42-band @xmath43 , morphology , @xmath44 color , petrosian radius , axis ratio , concentration index , color gradient in @xmath45 color , velocity dispersion , and equivalent width of the h@xmath46 line .
the @xmath42-band absolute magnitude @xmath43 is the ab magnitude converted from sdss magnitudes . to compute colors
, we use extinction ( schlegel et al .
1998 ) and @xmath47-corrected model magnitudes .
the superscript 0.1 means the rest - frame magnitude @xmath47-corrected to redshift of 0.1 ( blanton et al .
this makes galaxies at @xmath48 have @xmath47-correction of @xmath49 , independent of their spectral energy distributions ( seds ) .
all of our magnitudes and colors follow this convention , and the superscript will be subsequently dropped .
we also drop the @xmath50 term in the absolute magnitude .
we use the luminosity evolution correction of @xmath51 ( tegmark et al .
2004 ) .
we employ the @xmath52- and @xmath53-band atlas image and basic photometric parameters measured by the princeton / nyu group ] to measure the key physical parameters of the 317,533 galaxies in our sample .
the @xmath54 color gradient is given by the color difference between the region with @xmath55 and the annulus with @xmath56 , where @xmath57 is the petrosian radius .
we use the isophotal position angle , the seeing - corrected isophotal axis ratio in the @xmath53-band , and elliptical annuli in all parameter calculations to take into account flattening or inclination of galaxies .
the ( inverse ) concentration index , @xmath58 , is defined as @xmath59 where @xmath60 and @xmath61 are the radii from the center of a galaxy containing 50% and 90% of the petrosian flux in the @xmath53-band , respectively .
the @xmath54 color gradient and concentration index are corrected for the effects of seeing using sersic model fits of the @xmath52- and @xmath53-band images as described in paper i. we choose the concentration index as a measure of surface brightness profile because the sersic indices measured from sdss images are quite inaccurate except for large bright galaxies with @xmath62 , due to the relatively large seeing(@xmath63 ) of the sdss .
if the profile fit is made by avoiding the central part of galaxy images , which is much affected by the seeing , the seeing effects on the sersic index can be reduced but the fit becomes more unreliable due to the smaller fitting interval and noisier profile .
in contrast , by definition @xmath64 is insensitive to the surface brightness profile near the galaxy center within radius of @xmath60 .
our seeing - corrected concentration index seems accurate because it shows no dependence on either redshift or seeing .
we classify morphological types of galaxies using the prescription of park & choi ( 2005 ) .
galaxies are divided into early ( ellipticals and lenticulars ) and late ( spirals and irregulars ) types based on their locations in the @xmath65 versus @xmath54 color gradient space and also in the concentration index space .
the resulting morphological classification has completeness and reliability reaching 90% , as claimed by park & choi ( 2005 ) .
when photometry is excellent and galaxy images are well - resolved , more information from surface brightness fluctuations can be added for morphology classification .
but this is certainly not the case near the faint limit ( @xmath66 ) of the sdss sample we use . for the volume - limited sample d2
we perform an additional visual check of the color images of galaxies to correct misclassifications by the automated scheme for about @xmath67 blue early types ( those below the straight line in fig .
7 of paper i ) and red late types ( those with @xmath65 color redder than 2.4 ) .
the morphological types of 1.9% of galaxies , which are often blended or merged objects , are changed by this procedure .
we use petrosian radius to represent the physical size of galaxies .
the petrosian radii in our analysis , calculated by using elliptical annuli , are usually larger than those in the dr5 catalog , which adopted circular annuli .
the angular size is converted to a physical size in units of @xmath6kpc from the redshift assuming a @xmath68cdm universe with @xmath69 and @xmath70 .
the stellar velocity dispersion of sdss galaxies is measured by an automated pipeline called idlspec2d version 5 ( d. j. schlegel et al . 2007 , in preparation ) .
galaxy spectra of sdss galaxies are obtained by optical fibers with angular radius of @xmath71 .
the finite size of the sampled light smooths the central velocity dispersion profile . to correct the central velocity dispersion for this smoothing effects
we have adopted a simple aperture correction formula @xmath72 ( jrgensen et al .
1995 ; bernardi et al .
2003a ) , where @xmath73 is the equivalent circular effective radius @xmath74 and @xmath75 is the seeing - corrected effective angular radius in the @xmath53 band .
to find relations between intrinsic physical properties of galaxies and their environment we require a well defined and robust measure of environment .
we seek to use an environmental parameter that is defined directly from observational data , is continuous , and characterizes the full range of galaxy environments , from the most massive clusters to voids .
the local number density of galaxies at a given smoothing scale is such an example and will be the environmental parameter in our study .
local density gradient and shear can be also used to define environment , but will be considered in later studies .
our choice excludes parameters like the discrete environmental types such as cluster , group , or field .
assignment of galaxies to such environmental types are arbitrary to some extent and can be constructed based on continuous parameters when desirable .
local density estimation depends on the shape and size of the smoothing kernel .
we adopt the spline kernel with adaptive smoothing scale , @xmath76 to include a fixed number of galaxies within the kernel weighting .
this adaptive smoothing kernel is often used in smoothed particle hydrodynamics simulations .
we use the spline kernel because it is centrally weighted , unlike tophat or cylindrcal , and has a finite tail , unlike the gaussian .
the spline kernel was first used to estimate the local galaxy density by park et al .
( 1994 ) who demonstrated that the under - dense regions in the center for astrophysics ( cfa ) survey lack bright galaxies , indicating an environmental dependence of the luminosity function ( lf ) .
details of our method to calculate the local density using the spline kernel are summarized by c. park et al .
( 2007 , in preparation ) , who demonstrate that the smooth density field constructed by the spline kernel in redshift space recovers the real space density field much better than the cylindrical kernel at low- and intermediate - density regions .
the local density of galaxies sitting at very high density regions and with large peculiar velocities ( i.e. those appearing at the tips of fingers - of - god in redshift space ) is better measured by the cylindrical kernel , but the fraction of such galaxies is very small . because this smoothing kernel is adaptive , it has the important feature that it does not oversmooth dense regions , thereby blurring the `` walls '' into the `` voids . ''
cosmological simulations clearly show that the transition from the interior of large cosmic voids , which have nearly constant density , to the denser structures , defined by filaments and clusters , is quite abrupt .
maps of the observed redshift space distribution of galaxies support this picture ( see fig .
3 , below ) .
an adaptive smoothing kernel , such as the one we employ , better preserves that property of the galaxy distribution than a kernel with a fixed smoothing length or with an infinite tail . on the other hand , the adaptive nature of the smoothing kernel potentially complicates the interpretation , because the physical smoothing scale systematically varies with environment , from a small smoothing scale in clusters to a very large smoothing scale in voids .
however , the relationship between environment and smoothing scale is nearly monotonic and , unlike fixed - scale methods , the adaptive kernel has the advantage that the s / n for estimating density is more uniform .
we define the local density at a position in space as the spline kernel - weighted number density of @xmath17 galaxies surrounding that position . in every case
we use the volume - limited ls sample of @xmath17 galaxies with absolute magnitudes @xmath77 for local density estimation . in this way
the local density has a clear meaning , and it becomes possible to compare the local densities of galaxies in different volume - limited samples sitting at different radial distances .
the number of @xmath17 galaxies required within the smoothing volume is set to @xmath78 , which we find is the smallest number yielding good local density estimates . unless otherwise stated , @xmath78 ; below we also examine results with @xmath79 when studying larger scales . when searching for the spline radius containing the required number of galaxies , the smoothing volume can hit the sample boundaries or masked regions due to bright stars or galaxies .
in addition to this trouble , the angular selection function varies across the sky even within the survey region . we first check if the location in which the local density is being measured is within the survey area , and then ignore the survey geometry and angular selection function to search for the nearest 20 galaxies in the ls sample in order to find the spline radius and thus the local density .
we then randomly throw 1000 test particles into the spline radius around the location , and calculate the sum of normalized weights of the test particles assigned by the spline kernel , the angular selection function , and the radial boundaries .
the final local density is given by the first estimate divided by the sum of weights .
our results use only those local density estimates for which the correction factor is smaller than 2 .
the size of the spline kernel @xmath76 can be related with the effective gaussian smoothing scale using the fact that the effective smoothing volumes are @xmath80 and @xmath81 for spline and gaussian smoothings .
the result is @xmath82 . in the case of galaxies in sample d3
, we find the median values and 68% limits of @xmath83 are @xmath84 and @xmath85 or the effective @xmath86 and @xmath87 when @xmath88 and 200 are used , respectively .
the corresponding distributions of the local density and the effective gaussian smoothing length are shown in figure 2 .
the morphology - luminosity relation has been examined in paper i. in this paper we examine this relation as a function of local density environment , using our adaptive smoothing to label regions of space .
the upper panel of figure 3 shows the distribution of early ( red ) and late ( blue ) type galaxies in the d3 sample in a wedge defined by @xmath89 and @xmath90 in region 2 ( see fig .
1 ) of the sdss . the bottom panel shows the same galaxies , but distinguished by their absolute magnitudes ; bright galaxies with @xmath91 are red and fainter galaxies with @xmath92 are blue .
figure 4 shows distribution of galaxies in a slab with constant thickness of @xmath93mpc that cuts through region 2 .
overlaid are isodensity contours estimated from the local densities on a uniform grid through the middle of the slab .
the upper and lower panels use the local density calculated using the nearest @xmath94 and @xmath95 @xmath17 galaxies , respectively .
this comparison is made to study the dependence of environmental effects on scale in section 5 .
contours delineate isodensity regions of @xmath17 galaxies .
the thin lines mark the regions where the local density is @xmath96 or 0.1 for @xmath97 and @xmath98 , 2 , or 3 for @xmath99 , respectively .
the thick curves correspond to the local density equal to the mean density of the @xmath17 galaxies , which is measured to be @xmath100 .
we clearly see that early types prefer high - density regions , while late types are more uniformly distributed .
note carefully that the galaxies plotted are brighter than @xmath101 while the density field is determined by @xmath102 galaxies . throughout this paper
we use the ls sample to estimate local density , regardless of which volume - limited sample is being examined . figures 5 and 6 present the morphology - density - luminosity relation of sdss galaxies .
figure 5 shows the fraction of early type galaxies as a function of local density when the absolute magnitude is fixed .
we first remove late type galaxies in each sample whose seeing - corrected @xmath53-band isophotal axis ratio is @xmath103 in order reduce the internal absorption effect on absolute magnitude ( cf .
paper i ) .
we then bin galaxies in the absolute magnitude versus local density space .
the bin size in this two - dimensional parameter space is @xmath104 and @xmath105 .
we measure the fraction of early type galaxies , the median local density , and the median absolute magnitude of galaxies belonging to each bin , taking into account the fact that we have excluded some fraction of late type galaxies .
when we measure the fraction of early type galaxies , we need to know the number of early and late type galaxies . since we are only using late types with @xmath106 , we should infer the total number of late types given the number of those with @xmath106 .
some late types with @xmath103 must have dropped out of the volume - limited sample due to the internal absorption . in paper i
we showed that the correction factor can be reasonably made using the factor found in the faintest volume - limited sample of galaxies with @xmath107 .
the error bars are estimated from 16 random subsets of galaxies .
we find that the early type fraction is a monotonically increasing function of local density at fixed luminosity .
it is surprising to find that this monotonic dependence holds over 3 orders of magnitude in local density , from the highest down to the lowest density we explore , and in every luminosity bin .
the early type fraction falls well below 20% at extreme low densities .
this late type dominance at low density is particularly true for fainter galaxies . to test for the effects of the finite bin size in @xmath108 and @xmath43 on the relation we also examine the relation using smaller bin sizes .
the thin solid line between the lines of @xmath109 and @xmath110 is the morphology - density relation of galaxies in the @xmath111 sample with a median magnitude of @xmath112 using the bin sizes @xmath113 and @xmath114 .
the amplitude and slope of the relation is quite consistent with those for which larger bin sizes are used ; thus , our results are not affected by the finite bin sizes .
we also examine the effects of including those galaxies whose redshifts are borrowed from the nearest galaxy with redshift within @xmath115 , and find that the dependences of morphology fraction on density and luminosity are essentially the same .
our results on the morphology - density relation are consistent with , and extend to a larger range of density and luminosity , the findings of several earlier investigators .
for example , in studies limited to rich cluster environments , oemler ( 1974 ) and dressler ( 1980 ) found higher fractions of e and s0 types ( both classified as `` early type '' here ) in the centers of clusters .
subsequent investigations have verified and refined the understanding of this effect . as originally noted by postman & geller ( 1984 ) and by a number of more recent investigators ( lewis et al .
2002 ; goto et al . 2003 ; balogh et al .
2004a ; weinmann et al .
2006 ; quintero et al .
2007 ) this segregation is not limited to rich clusters , but occurs in poorer groups as well .
figure 5 reflects this fact in the strong dependence of the early type fraction on density even at the lowest densities .
the dependence of the early type fraction on absolute magnitude , even at fixed density , has been previously noted ( balogh et al .
2004b ; kauffmann et al .
2004 ; hogg et al .
as we will discuss later in more detail , the dependence of the late type subclass on density has also been seen as a slight dependence of the median blue sequence color on environment ( balogh et al . 2004b ; blanton et al . 2007 ) .
the main difference between our results here and these previous studies is that most of the previous studies use either visual classification or use simple classifications according to optical color .
although our classification system is related to broadband optical color , using the color gradients and concentrations yields a cleaner sample morphologically , bridging the gap between the studies based on visual classification and those based on purely color classification .
another important difference is the use of spline kernel to define local density .
the cylindrical kernel , which is frequently used in past studies ( goto et al .
2003 ) , is very noisy and can not distinguish among low density regions well while the local density estimated by the spline kernel holds a good tight correlation with the true density field down to extreme void environments ( c. park et al .
2007 , in preparation ) .
figure 6 presents the morphology - density - luminosity relation in a different way . here
we find that the early type fraction is a monotonically increasing function of luminosity at a given local density . in slightly over - dense regions , where the statistics are very good
, it can be clearly seen that the early type fraction rises abruptly at magnitudes brighter than @xmath116 ( see also fig . 9 of paper i ) .
the slope of the @xmath2-@xmath3 relation also has a break at a similar magnitude ( see fig .
14 , below ) .
this may imply that many early type galaxies more than about 1 mag brighter than @xmath117 have origins that are different from fainter ones .
it is remarkable that the early type fraction@xmath118absolute magnitude relation has similar slope for galaxies with local density ranging over nearly 3 orders of magnitude . at intermediate magnitudes , near @xmath117 , the early type fraction
is well described by the morphology - density - luminosity formula @xmath119(\rho/\bar{\rho})^{0.164}\ ] ] over the whole density range explored . at magnitudes fainter than about @xmath120
the local density dependence monotonically steepens at high density , but becomes weaker in under - dense regions ( see two bottom curves of fig .
a result of this is that in very high density regions the early type fraction decreases less steeply at fainter magnitudes than it does at lower densities ( see the top curve of fig . 6 ) .
balogh et al . ( 2004b ) measured the fraction of the red population of galaxies within fixed absolute magnitude bins as a function of projected local density @xmath121 adopted by balogh et al.(2004a ) .
even though galaxies are divided by a different criterion and the local density measure is different , the local density dependence they found is qualitatively quite similar . we emphasize that it is important to exclude highly inclined galaxies in this analysis
otherwise , the number of late type galaxies brighter than @xmath117 is significantly underestimated , and the slope of the early type fraction versus absolute magnitude relation is systematically biased high at bright magnitudes .
next we look for dependence of subclasses of late type galaxies on environment .
subclass classification of late type galaxies can be made in the color - color gradient space .
we divide the late type galaxies into two subclasses ; late late types with @xmath122 and early late types with @xmath123 .
this divides late type galaxies roughly into those of sc type and later and those earlier than sc , respectively ( park & choi 2005 ) .
the bin sizes are again @xmath124 and @xmath125 , and only the late types with axis ratio greater than @xmath126 are used . in figure 7
we find that the fraction of late spiral types with @xmath65 color bluer than @xmath127 increases strongly as the luminosity decreases .
this can be easily understood from the color - luminosity relation of late type galaxies ( see fig .
3@xmath128 of paper i ) .
this figure also shows that the late spiral population increases in low density regions for bright spirals , in particular .
even though our sub - spiral class division is very crude , it can be seen that the sub - type of spiral galaxies does depend on local density at fixed luminosity .
figure 8 shows the dependence of luminosity of galaxies in the d4 sample on the local density .
the upper panel is for the early types and the lower is for the late types with @xmath129 .
it is clear that high density regions contain very bright galaxies while low density regions do not .
this indicates that there are systematic effects of local density on the maximum luminosity that can be expressed by a change of luminosity function .
this environmental dependence of the luminosity function was found by park et al .
( 1994 ) in the cfa redshift survey data , and is now confirmed by the sdss data over 4 orders of magnitude of local density . also evident in figure 8 is that this local density dependence of luminosity is stronger for the early type galaxies .
the dependence is weaker but can be still observed for the late types .
the distribution of galaxies in the vertical direction at a given local density in figure 8 reflects the environment - dependent luminosity function which we plot in figure 9@xmath128 .
the top points and curve are the luminosity function measured from the d4 sample ( @xmath130 ) and its best - fit schechter function , respectively .
curves lying below are the luminosity functions of galaxies located in bins of different local density , normalized to the whole volume of d4 , so that the sum of these curves gives the total luminosity function .
we exclude late type galaxies with @xmath103 in the lf measurements to reduce the bias due to the internal absorption , and we weight the remaining late types by the inverse of their fraction .
the lf at the highest density ( _ lower solid line with open circles _ ) extends to magnitudes much brighter than those reached by the lf in the lowest density environment ( _ dotted line with triangles _ ) , in agreement our qualitative conclusion from figure 8 . to answer the question of whether or not the dependence of the lf of galaxies on local density
is due simply to the dependence of morphology fraction on local density or to a change of the shape of the lf , we normalize the lfs at the faint end and compare their shapes . in each density bin
used in figure 9@xmath128 we divide galaxies into early and late type groups , and recalculate their lfs .
figure 9@xmath131 and 9@xmath132 show that the shape of the lf of each morphological type does change as a function of local density .
in particular , the lf of early types is more sensitive to environment than that of late types .
therefore , the local density dependence of the lf is due both to the change of the early type fraction and that of the shape of the lf of each morphological type .
the visual results of figure 9 can be quantified by comparing the best - fit schechter function parameters .
we divide the galaxies in the d3 sample ( @xmath133 ) into late and early types and into seven different local density bins .
the solid lines with filled circles in figure 10 are the schechter function parameters @xmath117 ( _ upper panel _ ) and @xmath46 ( _ lower panel _ ) when the morphology subsets are combined .
we use the minuit package of the cern program library to estimate the parameters by the maximum likelihood method .
the estimated schechter function parameters are listed in table 2 .
the solid line with open circles and the dashed line with crosses in figure 10 are those of early and late types , respectively .
the luminosity of early type galaxies as measured by the characteristic absolute magnitude , @xmath117 , starts to rise steeply for local density above the mean density while that of late types rises below the mean density , stays more or less constant at intermediated density , and then rises again in high density regions .
previous work has also found that the high - luminosity turnover in the luminosity function strongly depends on local density or cluster richness ( e.g. , lumsden et al .
1997 ; driver et al . 1998
; marinoni et al . 2002 ; de propris et al . 2003 ; balogh et al . 2004a ; hoyle et al . 2005 ; popesso et al . 2005 ; zandivarez et al . 2006 ) .
croton et al .
( 2005 ) measured the schechter function parameters for early and late type galaxies in the 2dfgrs .
contrary to croton et al .
, we find that @xmath117 for early types is changing slowly below the mean density and that luminosity for late types rises at high density ( @xmath134 ) .
the difference seems to be due to difference in morphology classification .
they used a morphological classifier based on fiber spectra of galaxies .
morphology classifiers based simply on total color or spectrum tend to misclassify red spirals into early type , and blue ellipticals into late type .
the color gradient and concentration constraints in our morphological classification strongly reduce such misclassifications .
another important difference in our analysis is the exclusion of inclined ( @xmath103 ) late type galaxies .
when they are included , the number of relatively fainter galaxies is overestimated because many of the inclined galaxies are actually bright , but appear fainter due to internal absorption .
in particular , this systematic effect of internal absorption makes the @xmath46 parameter of late types more negative and makes the @xmath117 of late types appear fainter .
the rise of the characteristic luminosity of late types in high - density regions would not be detected unless the bright red spirals , which are abundant at high density , are correctly classified and the internal absorption effects are eliminated .
the faint - end slope , @xmath46 , does not show a monotonic behavior for late type galaxies , consistent with the results of hoyle et al .
( 2005 ) and croton et al .
it is interesting that @xmath46 for early types decreases ( steeper faint end slope ) at local density both below and above the mean . at local density near the cosmic mean
, the early type lf has an exceptionally flat faint end slope : @xmath135 .
detection of this behavior requires careful morphological classification , because faint early type galaxies reside predominantly not only in massive halos ( which prefer high density environment ) but also in low mass halos present in under - dense regions .
the latter tend to be fainter bluer early type galaxies that are correctly classified by our morphological classifier .
if galaxies were classified by color or their spectral properties alone , these blue early types would have been classified as late types .
we emphasize again that the major differences between our analysis and the previous ones lie in the morphological classification and exclusion of highly inclined late type galaxies , as well as the new local density estimator . on the other hand
, one should be cautious about the selection bias when the abundance of different species of galaxies are compared . in the case of the sdss catalog
there might be some bias due to the fact that some very low surface brightness galaxies are missed .
the schechter function parameters are measured from our volume - limited samples with magnitude limits of @xmath136 . in figure 9
we used the d4 sample ( @xmath137 ) to better see the behavior of the bright tail of the lf , and in figure 10 the d3 sample ( @xmath138 ) was adopted because this sample has the minimum uncertainties when both @xmath117 and @xmath46 are concerned .
however , these samples whose ranges in absolute magnitude are only about 1 mag below @xmath117 , are not really proper for giving accurate estimates of @xmath46 .
the limited range in absolute magnitude also develops some correlation between the estimates of @xmath117 and @xmath46 .
this is particularly so at low density , where @xmath117 is fainter , and for late types for which the faint end slope of the lf is steeper .
we compare the local density dependences of lf parameters measured from all five volume - limited samples with one another to see the effects of finite magnitude range . in figure 10 the results from the d2 sample ( @xmath139 )
are plotted as symbols without connecting lines for a comparison .
we confirm that the behaviors of @xmath117 of early and late type galaxies as a function of local density were basically the same for all samples .
it was also true for the @xmath46 parameter of early types .
however , the dependence of @xmath46 of late types on local density showed the systematic trend that @xmath46 became more negative as the faint limit of the sample became fainter . in order to measure @xmath46 of late types accurately
, one needs to use fainter galaxies , and for a flux - limited sample this makes one rely exclusively on the very close faint galaxies .
it should be noted that the nearby universe is a relatively under - dense part of the universe , and that the resulting parameters can be also biased due to the local density dependence of the lf .
dependence of @xmath46 on environments is still a matter of some debate ( e.g. , de propris et al .
1995 ; secker et al . 1997 ; valotto et al .
1997 , 2001 ; trentham 1998 ; andreon & cuillandre 2002 ; popesso et al .
2006 ) .
it is well known that galaxy colors are redder in regions of higher local density ( hogg et al . 2004 ; blanton et al .
2003c ; tanaka et al .
2004 ) . in this section
we show this trend is dominated by the dependence of color on luminosity and morphology . for this purpose
it was necessary to examine the dependence of color on local density when both morphology and luminosity are fixed .
we use @xmath65 color in this work because it is a measure of the star formation activity of galaxies in the recent past .
the top panels of figure 11 show scatter plots of @xmath65 color of early ( _ left panels _ ) and late ( _ right panels _ ) type galaxies in the d2 sample whose absolute magnitudes are between @xmath140 and @xmath141 ( the median magnitude is @xmath142 ) .
the thick blue solid lines delineate the most probable ( mode of the distribution of color at fixed absolute magnitude and at fixed local density interval ) color of early types and the median color of late types .
late type galaxies with axis ratios @xmath143 are again excluded .
the short - dashed green lines are the most probable ( early ) and median ( late ) colors of galaxies with @xmath144 .
the corresponding scatter plots are not shown to avoid confusion .
the long - dashed magenta line and the red solid line are for galaxies in magnitude bins of @xmath145 and @xmath146 with median magnitudes of @xmath110 , and @xmath147 , respectively .
we first note that there is no break in color as a function of local density for either early or late type galaxies .
tanaka et al . ( 2004 ) reported detection of a break in color at a density of @xmath148 galaxies @xmath149mpc@xmath150 ( see tanaka et al . for the definition of projected local density @xmath121 ) for galaxies with magnitudes @xmath151 .
it is clear that this result is simply due to the fact that blue late types tend to avoid in very high density environments , and it is the morphology bias rather than the color bias that causes the effect .
this feature in the morphology bias appears as a break in the morphology - luminosity relation at an absolute magnitude of about @xmath152 ( paper i ) .
second , we note that the color of galaxies is a very insensitive function of local density once the morphology and luminosity are fixed . as the local density varies from @xmath153 to about 0.37 , the color changes only by 0.03 for early types and by 0.11 for late types ; both get redder at higher density .
this behavior is insensitive to luminosity , which can be seen from the fact that the lines are nearly parallel to one another . on the other hand , the color changes by 0.22 and 0.75 as the magnitude changes by 2.2 from @xmath154 to @xmath147 for early and late types , respectively .
the difference between the typical colors of early and late types is as large as 1 mag when @xmath155 .
therefore , the dependence of color on local density is dominated by the dependences of color on morphology and luminosity , which in turn are strongly dependent on local density , as shown in sections 4.1 and 4.2 .
figure 11 also shows that blue early types live mainly at intermediate- and low - density environments , and that some red early type galaxies still exist in the extreme low density regions with @xmath156 .
tanaka et al . ( 2004 ) reported that the star formation rate indicated by the @xmath54 color and the equivalent width of the @xmath157 emission line is suppressed in dense environments .
it turns out that most of the effects they detected are due to the dependence of morphology and luminosity on local density . to inspect the collective behavior of color of each morphological type we examine the early and late type sequences in the color - magnitude diagram .
the points in figure 12 are early ( _ left panels _ ) and late ( _ right panels _ ) type galaxies in the d3 sample located in high ( _ upper panels _ ) and low ( _ bottom panels _ ) density environments . in the upper panels ,
solid lines are the most probable color for the early types and the median color for the late types . in the bottom panels ,
the most probable and median color curves at high density are replotted as dashed lines , for comparison with the curves at lower density .
we find that the shift of the early type sequence in the color - magnitude diagram toward bluer color is only 0.025 mag when the local density changes from @xmath158 to @xmath159
. a larger shift of 0.14 in @xmath65 color is observed for the late type sequence .
the very slight dependence of the location of the red sequence on the density field has been noted in the past , originally by sandage & visvanathan ( 1978 ) , but more recently by others ( balogh et al . 2004b ; hogg et al .
2004 ; bernardi et al .
2006 ; blanton et al .
the stronger dependence of the blue sequence colors on environment was pointed out by balogh et al .
( 2004a ) .
the points in figure 11@xmath131 are the color gradient of galaxies in the d2 sample as a function of local density .
the most probable color gradient of early type galaxies is slightly negative ( the center is redder ) and shows no environmental dependence at a given absolute magnitude .
it is nearly independent of luminosity as well .
this extraordinary result is due to the fact that the internal color gradient of early type galaxies is almost independent of luminosity even though the color itself does have a monotonic dependence on luminosity ( cf .
fig . 3@xmath128 and 3@xmath131 of paper i ) .
late type galaxies have more negative color gradients when they are brighter ( cf .
3@xmath131 of paper i ) , which explains the dependence of the color gradient of late types on absolute magnitude in the right panel of figure 11b .
this may reflect a change in the bulge - to - disk ratio with luminosity .
figure 11@xmath128 shows that fainter late type galaxies are bluer in their integrated color than bright ones , and figure 11@xmath131 indicates that they have bluer centers ( thus lower color gradients ) .
the color difference weakly depends on local density for faint late type galaxies while no dependence is seen for bright galaxies . at fixed color or
estimated star - formation history , several previous authors found little dependence of concentration on environment ( hashimoto & oemler 1999 ; bernardi et al .
2003a ; kauffmann et al .
2004 ; blanton et al . 2005a ; quintero et al .
we extend these results by showing that even when we use a cleaner definition of morphology , the radial profiles of galaxies appear to be only weakly related to their environment . when morphological type and luminosity are fixed , the surface brightness profile , represented by the ( inverse ) concentration index , is nearly independent of local density as shown in figure 11@xmath132 .
the median @xmath58 of bright early types with absolute magnitudes within 0.8 mag bins shows little change at all local densities explored .
therefore , the dependence of surface brightness profile of early types on local density is solely due to the dependence of the profile on luminosity ( cf . fig .
3c of paper i ) , which monotonically depends on local density , as demonstrated in section 4.2 .
faint early types seem to be slightly more concentrated at low density .
it is difficult to see a dependence of concentration index of late types on local density because the dispersion in @xmath58 is large and the effects seem weak .
the size ( petrosian radius ) of galaxies is an increasing function of luminosity as shown in figure 4@xmath128 of paper i. figure 13@xmath128 shows that physical size is almost independent of local density when luminosity and morphology are fixed .
we detect a small monotonic dependence of galaxy size on local density for both early and late types , except for the very bright galaxies .
surprisingly , galaxies of the same luminosity are smaller and therefore have higher surface brightness at low density .
galaxies with @xmath160 located at low density ( @xmath161 ) are about 8% smaller in size than those at high density ( @xmath162 ) , for both early and late types . the weak dependence of size on local density
should be accepted with caution , because including light from neighboring galaxies can in principle create a trend in the sizes with density in the direction we see .
kauffmann et al . ( 2004 ) and blanton et al . ( 2005a ) have both investigated the relationship between galaxy size and environment ( at fixed luminosity and at fixed star - formation history or color , respectively ) , and both found only small effects . in the range of luminosities we consider here , more luminous than @xmath117 , blanton et al .
( 2005a ) find a trend for larger galaxies to be in denser regions as well ( although they found this trend is reversed at lower luminosities ) .
bernardi et al .
( 2003b ) investigated the dependence of residuals of early type galaxies from the fundamental plane on environment and find that the residuals from the fundamental plane in effective surface brightness correlate slightly with environment .
the sense is that at fixed luminosity and velocity dispersion , galaxies are larger in dense regions , similar to our results here . in the right panel of figure 13@xmath131
we see that the central velocity dispersion of late type galaxies shows no dependence on local density . however , we find an interesting dependence of the @xmath2-@xmath3 relation on environment for early type galaxies .
the velocity dispersion of fainter early types slightly increases toward low - density regions , while the opposite is true for brighter early types .
this might be evidence for an environment - dependent faber - jackson relation or , more generally , such a dependence of the @xmath2-@xmath3 relation . to explore this possibility further we plot early type galaxies in high and low - density regions in the absolute magnitude versus velocity dispersion space .
the left panels of figure 14 show the scatter plots of early type galaxies in the d4 sample located at different density and their most probable relations in the velocity dispersion - magnitude diagram .
the dashed curves in the lower panel are equal to the solid ones in the upper panel .
we clearly see that relatively bright early type galaxies in higher density regions have higher central velocity dispersions than those in under - dense regions .
the right panels are for the galaxies in the d2 sample and show the relations down to fainter magnitudes .
these plots indicate that the slope of the @xmath2-@xmath3 relation depends on local density and that this trend is more evident at faint magnitudes .
if we fit a power - law relation @xmath163 to the curves of most probable velocity dispersion , we measure @xmath164 and @xmath165 in regions with local densities of @xmath166 and @xmath159 , respectively , near @xmath167 .
thus , not only is the relation between luminosity and @xmath3 not well represented by a single power - law ( cf . fig .
4@xmath131 of paper i ) , the best - fitting such relation depends on local density of the galaxies .
the equivalent width of the @xmath168 line is measured from the sdss spectra , and is often used as a measure of recent star formation activity ( or nuclear activity ) . because sdss spectra are obtained from the central @xmath71 radius region of galaxies ,
this line width is not necessarily a representative measure of star formation activity , particularly for spirals with a significant bulge component .
nevertheless , the star formation activity as measured by the equivalent width of the @xmath168 line , is on average stronger for faint late type galaxies ( cf .
4@xmath132 of paper i ) .
we find that @xmath157 line strength varies little with local density once the morphology and luminosity are fixed .
note that a strong local density dependence of @xmath169 can be obtained if all morphological type and luminosity are mixed because the fraction of late type and faint galaxies increases at lower densities .
however , we are able to detect a weak correlation between @xmath157 line width and local density for late type galaxies ; the star formation activity signaled by the @xmath157 line strength is relatively stronger at low density environment at a fixed luminosity
. for galaxies with @xmath170 a linear fit to the median relation gives log@xmath171 .
this dependence is weaker for brighter galaxies .
this variation of star formation activity with environment is consistent with the findings of rojas et al.(2005 ) , who compared star formation activity of galaxies in voids with that of galaxies of similar luminosity in denser environments .
other investigators who have explored the dependence of h@xmath46 emission on density have found that even at fixed galaxy concentration there is a strong relationship between the fraction of emission line galaxies and density
. however , gmez et al . ( 2003 ) and balogh et al . ( 2004a ) found , as we do here , that the distribution of h@xmath46 equivalent widths for emission line galaxies are only weakly related to density .
these results parallel the similarly weak relationship between color and environment for late type galaxies .
the majority of early type galaxies show vanishing star formation activity regardless of luminosity .
we note that there is a group of active early types located at median or low density environments .
the points in figure 15 are the isophotal @xmath172 ratio of early type galaxies in the d4 sample with absolute magnitudes between @xmath173 and @xmath174 as a function of local density .
their median relation is the thick solid line .
other lines are for brighter galaxies in the same sample , and the corresponding scatter plots are not made , in order to avoid confusion .
brighter early types are rounder than fainter ones ( see also fig . 5 of paper i ) , which might be explained by the tidal interaction theory ( cf .
thuan & gott 1977 ) .
once the luminosity is fixed , the local density dependence of the axis ratio of early types is so weak that it is not detected in this plot .
the bottom panel of figure 15 compares the histograms of the @xmath172 ratio of galaxies at very high and very low density environments .
no statistically significant difference is observed .
so far , we have used local density estimated by the spline kernel smoothing in which the kernel size varies to include @xmath175 galaxies in the ls sample of @xmath176 galaxies .
we now inspect the environmental dependence of physical properties of galaxies when the smoothing kernel includes 200 galaxies . because we have found that most galaxy properties are nearly independent of local density once morphology and luminosity are fixed , we look for the possible dependence of only the morphological fraction and luminosity function on the much larger scale density field .
the isodensity contours of the upper and lower plots in figure 4 compare the two density fields @xmath177 and @xmath178 smoothed over @xmath175 and @xmath95 , respectively .
figure 16 shows the early type galaxy fraction as a function of @xmath179 at each fixed luminosity with an absolute magnitude bin size of 0.8 .
surprisingly , the nature of the relations is basically the same as that found by using a smaller smoothing scale of @xmath175 ( although the relation seems steeper for fainter galaxies ) .
we have checked that we reach the same conclusion even for @xmath180 .
figure 17 demonstrates the dependence of lf of galaxies on @xmath179 .
the schechter function parameters @xmath117 and @xmath46 depend on the larger scale density @xmath181 in the same way that they do on the smaller scale density @xmath182(see fig .
the characteristic luminosity strongly increases as @xmath181 rises .
the type - specific parameters also behave in the same way .
the only noticeable difference is the @xmath46 parameter of the combined sample at the lowest density , but the statistical significance of that single point is low . it should be noted that the smoothing scale corresponding to @xmath183 is not far from the linear scale ( the rms density fluctuations is about @xmath184 while it is @xmath127 for @xmath175 ) .
the morphology and luminosity of galaxies appear to respond sensitively to the linear density field . since the density field that has @xmath185 smoothing
is correlated with that having @xmath88 , it is not surprising to find similar relations between the local density and galaxy properties at both smoothing scales . therefore , it is now very important to identify the scale on which the large - scale environmental dependence of morphology and luminosity of galaxies originates . to answer this question we study the dependence of the morphological fraction on very _ small _ scale environment by extracting samples both with and without close neighbors .
we select `` isolated '' galaxies brighter than @xmath186 in the d2 sample d2 ( @xmath187 ) .
we adopt the isolation criteria for a target galaxy with @xmath43 that there can be no companion galaxy in the absolute magnitude interval @xmath188 , with redshift difference within 300 km s@xmath189 , and with separation across the line of sight less than @xmath190kpc for a candidate galaxy @xmath53 .
the solid line in figure 18 is the early type fraction of isolated galaxies as a function of @xmath191 .
since the morphological fraction of isolated galaxies with no close companion still correlates tightly with @xmath182 , we conclude that the large - scale environment is definitely one of the factors that determine the morphology of galaxies .
we reach the same conclusion when the isolation criteria are changed to 500 km s@xmath189 in redshift difference and @xmath192 mpc in the separation across the line of sight . in figure 18
the long and short dashed lines are the early type fractions of the galaxies with only one neighbor ( binary ) or those with one or more companions ( unisolated ) , respectively .
we see that galaxies having close companions are more likely to be of early type than isolated galaxies , even at the same large - scale density .
this demonstrates that the morphology of galaxies is determined in response to very small scale density as well as to large ( @xmath175 ) and very large ( @xmath183 ) scale background density fields . to narrow down the scale responsible for the origin of morphology we push our study further using the densities at these three scales in figures 19 and 20 .
figure 19 shows the relation between the densities estimated with @xmath88 and 200 at each galaxy in the d2 sample .
superposed are contours of constant early type galaxy fractions of 0.25 , 0.30 , 0.35 , 0.40 , and 0.45 from left to right .
we note that the contours are basically vertical .
the result shows that the morphology of galaxies sitting at a fixed @xmath177 is independent of the larger scale density @xmath178 and that the dependence of morphology fraction on the density with @xmath185 smoothing is merely because of its correlation with the density when @xmath88 .
therefore , the origin of galaxy morphology must be due to the environmental conditions at scales much lower than the @xmath185 scale , which is roughly the @xmath193 mpc gaussian smoothing scale .
figure 20 compares @xmath177 with the distance to the nearest neighbor from a galaxy , @xmath194 , representing the small - scale density in this logarithmic plot . for each galaxy in sample d2 ( @xmath195 ) that has absolute magnitude limits @xmath196 we find the nearest companion that lies in the range @xmath197 and has radial velocity difference less than 500 km s@xmath189 , which is a little more generous than that used above .
if luminosity is proportional to mass , 1.5 mag fainter luminosity corresponds to 25% of mass .
our criterion implies that a companion with mass greater than 25% of a target galaxy s mass is heavy enough to influence the morphology of the galaxy .
superposed are again contours of constant early type fractions of 0.32 , 0.35 , 0.40 , 0.50 , and 0.55 .
it shows interesting dependences of morphology fraction on both scales .
in general , the contours have positive slopes , which means the early type fraction increases as @xmath194 decreases at fixed @xmath177 . at fixed @xmath194 the fraction increases as @xmath177 increases , which is already shown in figure 18 for @xmath198mpc .
the strongest dependence on @xmath194 is observed when @xmath199 mpc .
the dependence on @xmath194 is weaker when @xmath194 is higher or lower . in under - dense regions below @xmath200 the majority of galaxies
are isolated , with @xmath201mpc , and the morphology fraction is much less dependent on the existence of neighboring galaxies . in this under - dense environment galaxies are mostly of late type and morphology is sensitive to @xmath177 .
the slope of contours depends on our choice of companion criteria and the smoothing size used to calculate the morphology fraction .
but the general trend described above remains robust .
a previous study analogous to ours is balogh et al .
( 2004b ) who claimed that the fraction of emission line galaxies depends on both @xmath202 and @xmath203 mpc gaussian smoothing scales .
importantly , they gave some evidence that the fraction was mostly dependent on local density @xmath204 where @xmath205 is high , but depended more on background density @xmath205 where @xmath205 is lower .
their claim is in agreement with our findings ( compare their fig . 9 with fig .
20 of this paper ) .
lccccc all & all & @xmath206 & @xmath207 & @xmath208 & @xmath209 + & @xmath210 & @xmath211 & @xmath212 & @xmath213 & @xmath214 + & @xmath215 & @xmath216 & @xmath217 & @xmath218 & @xmath219 + & @xmath220 & @xmath221 & @xmath222 & @xmath223 & @xmath224 + & @xmath225 & @xmath226 & @xmath227 & @xmath228 & @xmath229 + early & all & @xmath230 & @xmath231 & @xmath232&@xmath233 + & @xmath234&@xmath235 & @xmath236 & @xmath237&@xmath238 + & @xmath239&@xmath240 & @xmath241 & @xmath242&@xmath243 + & @xmath244&@xmath245 & @xmath246 & @xmath247&@xmath248 + & @xmath249&@xmath250 & @xmath251 & @xmath252&@xmath253 + late & all & @xmath254&@xmath255&@xmath256&@xmath257 + & @xmath210&@xmath258&@xmath259&@xmath260&@xmath261 + & @xmath215&@xmath262&@xmath263&@xmath264&@xmath265 + & @xmath220&@xmath266&@xmath267&@xmath268&@xmath269 + & @xmath270&@xmath271&@xmath272&@xmath273&@xmath274 +
we have studied dependences of physical properties of galaxies in the sloan digital sky survey ( sdss ) on environment .
there are several differences between our analysis and previous studies .
first , we have used smooth local density estimated by the spline kernel enclosing 20 or 200 @xmath17 galaxies to define environment .
because the kernel used is centrally weighted , finite , and variable in size , it gives us a wide dynamic range in the estimated local density over 4 orders of magnitude .
second , we have used an accurate automated morphology classifier to group large numbers of galaxies into early and late types . in the case of the d2 sample with minor correction by visual inspection
we expect the completeness and reliability to exceed 93% .
these high quality morphology sets significantly reduce the biases due to misclassification .
third , we have used a set of volume - limited samples defined by absolute magnitude .
environmental dependences are studied for these volume - limited samples at fixed morphology and luminosity . in this way we have avoided confusion due to mixed dependences that occur when
flux - limited samples are used .
we have also attempted to reduce the effects due to internal extinction by excluding late type galaxies with axis ratios less than @xmath275 .
thanks to the large size of the sdss spectroscopic catalog of galaxies , we are able to measure the environmental dependences to reasonable accuracies even after dividing galaxies into many subsamples selected by morphology and luminosity .
our major findings are as follows .
the fraction of early morphological type galaxies is a monotonically increasing function of a ( well - defined ) local density and luminosity .
the morphology - density - luminosity relation , as measured in this work , should be a key constraint on galaxy formation models .
it remains for us to understand what causes the morphology of a galaxy at fixed local density .
we find that morphological fractions vary strongly with local density , but we find a mix of both early and late types at all densities .
thus , conditions other than local density must contribute to determining the morphology of a galaxy at fixed local density .
the extra cause might be environmental properties other than density .
the gravitational potential shear field is an example of such an environmental property .
morphology can also depend on the merger history of each object , which can be stochastic to some degree .
we find that the environmental dependences of various galaxy properties are almost entirely due to the dependences of morphology and luminosity on environment .
when morphology and luminosity are fixed , other physical properties , such as color gradient , concentration , size , velocity dispersion , @xmath168 emission , and axis ratio , are nearly independent of local density .
careful inspection reveals weak residual local density dependences for several properties .
we find that the @xmath65 color of late type galaxies is monotonically bluer at low density environments , even at a fixed luminosity .
early types show much weaker dependence .
we also find that bright early type galaxies in high density regions have higher central velocity dispersions compared to those in low density regions , but that the opposite is true for fainter galaxies .
a consequence is a local density dependence of the @xmath2-@xmath3 relation .
because galaxy properties strongly depend not only on luminosity but also on morphology , it is clear that galaxy properties can not be determined solely by dark halo mass . our results , using a new morphological classifier and a robust method of local density estimation , are largely consistent with several previous studies attempts ( kauffmann et al .
2004 ; blanton et al . 2005a ; hashimoto & oemler 1999 ; christlein & zabludoff 2005 ) . in each of those papers
, the authors found a relationship between galaxy color and environment at fixed concentration , a measure of galaxy structure associated with morphology ( strateva et al .
conversely , those papers showed that at fixed color there is no relationship between concentration and environment .
those results regarding concentration do not refer to morphology as we define it here , which indeed includes @xmath65 color as one of key indicators .
in park & choi ( 2005 ) we have shown that morphology can be traced not simply by concentration alone , but by a combination of concentration , color , and color gradient .
once morphology is defined this way , almost any other indicator of type appears to be nearly fixed with environment , even as the fraction of galaxies of each morphology changes .
indeed , they imply that because the morphological mix changes with local density , at fixed concentration one should indeed measure a correlation between color and density .
the converse result , that concentration shows only a weak relationship with density at fixed color , is a consequence of the fact that color is the more powerful type indicator .
we show that morphology of galaxies appears to depend on local density , from very small scales ( @xmath276 mpc ) up to the @xmath78 and 200 smoothing scales corresponding to about 4.7 and @xmath4 mpc effective gaussian smoothing scales .
it is shown that the environment at scales larger than about @xmath4 mpc do not independently affect galaxy morphology .
this gives us an important clue to understanding the origin of galaxy morphology .
the direct cause of galaxy morphology is not related with the background density of the linear - regime scale at the present epoch .
we have also shown that morphology of bright galaxies depends on both the local density with @xmath78 smoothing and the distance to the nearest companion galaxy .
the dependence on @xmath194 becomes strongest when the neighbor distance is about @xmath277 kpc at which the combined effects of strong tidal force and long duration of interaction on galaxy property seems to reach the maximum .
the importance of the smooth density with @xmath78 rises at larger and smaller separations . in the currently popular scenario of galaxy formation and evolution
the properties of galaxies depend either only on the formation history of their dark matter halos or also on nurture processes such as ram pressure stripping , strangulation , and harassment .
in the cold dark matter ( cdm ) cosmogony , a galaxy - scale dark halo forms through a series of mergers of subgalactic objects .
the over - density at the galaxy scale , which is statistically related with the larger scale density , will determine the final mass .
other physical properties of the galaxy will depend on the final mass , but much variation is expected , depending on how the final mass is actually assembled .
gott & thuan ( 1976 ) proposed that galaxy morphology is determined by the amount of gas left over at maximum collapse of the protogalaxy , which in turn depends on the ratio of the star formation time scale @xmath278 to the collapse time @xmath279 .
they claimed that ellipticals had relatively small @xmath280 ( if a @xmath281 star formation law was adopted ) and had finished star formation by the time of maximum collapse .
in additional to this initial condition and sub - galactic - scale evolution , the local environment can also directly affect galaxy properties .
ram pressure stripping ( gunn & gott 1972 ) is a mechanism that removes gas and shuts off star formation when a galaxy orbits a hot cluster halo .
the morphology of a spiral galaxy can be transformed to an early type by this process .
high - speed encounters of galaxies with other halos typically in clusters cause impulsive heatings , called harassment ( moore et al . 1996 ) , and can transform spirals to early types .
another mechanism , strangulation ( balogh et al .
2000 ) , can also transform morphology through a decline of star formation rate due to shut - off of the newly accreted gas when a galaxy enters a cluster or group environment and loses its hot gas reservoir .
there is also observational and theoretical evidence that the tidal force field in clusters can transform the infalling spirals to early types ( moss & whittle 2000 ; gnedin 2003 ) .
it is important to note that all these nurture mechanisms proposed so far are basically effective in group or cluster environments . in previous studies
it was claimed that galaxy properties such as star formation activity and morphology fraction showed discontinuities at a critical density for galaxies somewhat fainter than the @xmath17 galaxy .
this finding was used to support the proposition that properties of the relatively faint galaxies are affected by morphology transformation mechanisms working in the group or cluster environment ( tanaka et al .
2004 ; balogh et al .
2004 ) . however , we find that there is no feature in the galaxy property versus local density relation even for such relatively faint galaxies when luminosity and morphology are fixed .
the only feature is the sharp decrease of the late - type fraction above the critical luminosity of about @xmath1 in the morphology versus luminosity relation ( paper i ) .
this indicates that there seems to be morphology and luminosity - determining processes that work at all local densities in a continuous way .
the strong effect of a close companion on galaxy morphology as indicated by this study , might imply the existence of a morphology transformation mechanism acting at a distance without a direct physical contact .
what is important is that this process seems to be working in a wide density environment .
figure 20 indicates that this effect is maximal for galaxies having a companion at a separation of about @xmath282 kpc .
this might be because the tidal effects are strong enough to significantly enhance the star formation rate , and the duration of interaction is large as well at this particular separation . the morphology transformation process caused by this rather secular effects in binary or few - body systems
can be smoothly connected with the more instantaneous tidal interactions with the cluster core or local substructures in clusters ( moss & whittle 2000 ; gnedin 2003 ) .
galaxies can be either isolated or in binary / multiple systems at all environments .
for example , in figure 18 the fraction of unisolated galaxies is about 20% even in the low density region with @xmath283 . a similar finding
was reported in the existence of red galaxies in low density regions ( balogh et al .
figure 18 also shows that the early type fraction is systematically higher for galaxies located at higher @xmath284 no matter whether they are isolated or in systems .
the isolated early types may have been initially born as early types or may be final merger products of initially multiple systems .
it is reasonable to believe that our local background density @xmath182 calculated by using neighboring 20 @xmath17 galaxies is a good indicator of the ratio @xmath280 at the time of galaxy formation . the merger rate during the subsequent evolution and the average rates of all the nurture processes mentioned above are likely to be increasing functions of @xmath182 .
the fraction of galaxies in binary / multiple systems are also an increasing function of @xmath182 .
if these are true and if small @xmath280 or above evolution processes yield early type galaxies , we can understand the overall dependence of morphology on @xmath182 .
the increase of the early type fraction in binary or multiple systems can be explained by the tidal disturbance effects .
it remains to be understood why the dependence on the nearest companion distance becomes maximum near the particular separation of @xmath277 kpc , which is roughly the separation between two typical bright galaxies when their dark halos are just touching .
the authors thank chan - gyung park for measuring the luminosity function and yeong - shang loh for helpful comments .
cbp acknowledges the support of the korea science and engineering foundation ( kosef ) through the astrophysical research center for the structure and evolution of the cosmos ( arcsec ) .
msv acknowledges support from nasa grant nag-12243 and nsf grant ast-0507463 .
msv thanks the department of astrophysical sciences at princeton university for its hospitality during sabbatical leave .
yyc , cbp , and msv thank the aspen center for physics , at which much of this paper was written .
jrg is supported by nsf grant ast 04 - 06713 .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , cambridge university , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
, abazajian , k. , et al .
2004 , , 128 , 502 , adelman - mccarthy , j. k. , et al .
2007 , submitted , andreon , s. , & cuillandre , j .- c .
2002 , , 569 , 144 , balogh , m. , et al .
2004a , , 348 , 1355 , balogh , m. l. , baldry , i. k. , nichol , r. , miller , c. , bower , r. , & glazebrook , k. 2004b , , 615 , l101 , bernardi , m. , nichol , r. c. , sheth , r. k. , miller , c. j. , & brinkmann , j. 2006 , , 131 , 1288 , bernardi , m. , et al .
2003a , , 125 , 1817 , bernardi , m. , et al .
2003b , , 125 , 1866 , blanton , m. r. , & berlind , a. a. 2007 , in press ( astro - ph/0608353 ) , blanton , m. r. , eisenstein , d. , hogg , d. w. , schlegel , d. j. , & brinkmann , j. 2005a , , 629 , 143 , blanton , m. r. , et al .
2005b , , 129 , 2562 , blanton , m. r. , lin , h. , lupton , r. h. , maley , f. m. , young , n. , zehavi , i. , & loveday , j. 2003a , , 125 , 2276 , blanton , m. r. , et al . 2003b , , 125 , 2348 , blanton , m. r. , et al .
2003c , , 594 , 186 , boselli , a. , & gavazzi , g. 2006 , , 118 , 517 , choi , y .- y . , park , c. , & vogeley , m. s. 2006 , , submitted ( paper i ) , christlein , d. , & zabludoff , a. i. 2005 , , 621 , 201 , colless , m. , et al .
2001 , , 328 , 1039 , croton , d. j. , et al .
2005 , , 356 , 1155 , de propris , r. , pritchet , c. j. , harris , w. e. , & mcclure , r. d. 1995 , , 450 , 534 , de propris , r. , et al .
2003 , , 342 , 725 , dressler , a. 1980 , , 236 , 351 , driver , s. p. , couch , w. j. , & phillipps , s. 1998 , , 301 , 369 , fukugita , m. , ichikawa , t. , gunn , j. e. , doi , m. , shimasaku , k. , & schneider , d. p. 1996
, , 111 , 1748 , gnedin , o. , y. 2003 , , 589 , 752 , gmez , p. , et al .
2003 , , 584 , 210 , goto , t. , et al .
2003 , , 346 , 601 , gott , j. r. , & thuan , t. x. 1976 , , 204 , 649 , gunn , j. e. , et al .
2006 , , 131 , 2332 , gunn , j. e. , et al .
1998 , , 116 , 3040 , gunn , j. e. , & gott , j. r. 1972 , , 176 , 1 , hashimoto , y. & oemler , a. j. 1999 , , 510 , 609 , hamilton , a. j. s. & tegmark , m. 2004 , , 349 , 115 , hogg , d. w. , et al .
2004 , , 601 , l29 , hogg , d. w. , finkbeiner , d. p. , schlegel , d. j. , & gunn , j. e. 2001 , , 122 , 2129 , hoyle , f. , rojas , r. r. , vogeley , m. s. , & brinkmann , j. 2005 , , 620 , 618 , ivezi , z. , et al .
2004 , an , 325 , 583 , jrgensen , i. , franx , m. , & kjrgaard , 1995 , , 276 , 1341 , kauffmann , g. , white , s. d. m. , heckman , t. m. , m ' enard , b. , brinchmann , j. , charlot , s. , tremonti , c. , & brinkmann , j. 2004 , , 314 , lewis , i. , et al .
2002 , , 334 , 673 , lumsden , s. l. , collins , c. a. , nichol , r. c. , eke , v. r. , & guzzo , l. 1997 , , 290 , 119 , lupton , r. h. , gunn , j. e. , ivezc , z. , knapp , g. r. , kent , s. , & yasuda , n. 2001 , in asp conf .
238 , astronomical data analysis software and systems x , ed .
f. r. harnden , jr .
, f. a. primini , & h. e. payne ( san francisco : asp ) , 269 , marinoni , c. , hudson , m. j. , & giuricin , g. 2002 , , 569 , 91 , moss , c. , whittle , m. 2000 , , 317 , 667 , oemler , a. 1974 , , 194 , 1 , park , c. , choi , y .- y . , & kim , j. 2006 , in preparation , park , c. , et al .
2005 , , 633 , 11 , park , c. , & choi , y .- y . , 2005 , , 635 , l29 , park , c. , vogeley , m. s. , geller , j. , & huchra , j. p. 1994
, , 431,569 , pier , j. r. , munn , j. a. , hindsley , r. b. , hennessy , g. s. , kent , si . m. , lupton , r. h. , & ivezc , r. 2003 , , 125 , 1559 , popesso , p. , b " ohringer , h. , romaniello , m. , & voges , w. 2005 , , 433 , 415 , popesso , p. , biviano , a. , bhringer , h. , & romaniello , m. 2006 , , 445 , 29 , postman , m. , & geller , m. 1984 , , 281 , 95 , quintero , a. d. , berlind , a. a. , blanton , m. r. , & hogg , d. w. 2006 , , submitted ( astro - ph/0512004 ) , rojas , r. r. , vogeley , m. s. , hoyle , f. , & brinkmann , j. 2004 , , 617 , 50 , rojas , r. r. , vogeley , m. s. , hoyle , f. , & brinkmann , j. 2005 , , 624 , 571 , schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 , sandage , a. , & visvanathan , n. 1978 , , 225 , 742 , secker , j. , harris , w. e. , & plummer , j. d. 1997 , , 109 , 1377 , smith , j. a. , et al .
2002 , , 123 , 2121 , stoughton , c. , et al .
2002 , , 123 , 485 , strateva , i. , et al .
2001 , , 122 , 1861 , strauss , m. a. , et al .
2002 , , 124 , 1810 , tanaka , m. , et al .
2004 , , 128 , 2677 , tegmark , m. , et al .
2004 , , 606,702 , trentham , n. 1998 , , 293 , 71 , thuan , t. x. , & gott , j. r. 1977 , , 216 , 194 , tucker , d. , et al .
2006 , an , in press , valotto , c. a. , moore , b. , & lambas , d. g. 2001 , , 546 , 157 , valotto , c. a. , nicotra , m. a. , muriel , h. , & lambas , d. g. 1997 , , 479 , 90 , weinmann , s. m. , van den bosch , f. c. , yang , x. , & mo , h. j. 2006 , , 366 , 2 , york , d. , et al .
2000 , , 120 , 1579 , zandivarez , a. , martnez , h. j. , & merchn , m. e. 2006 , astro - ph/0602405 | we investigate the dependence of physical properties of galaxies brighter than @xmath0 in the sloan digital sky survey ( sdss ) on environment , as measured by local density using an adaptive smoothing kernel .
we find that variations of galaxy properties with environment are almost entirely due to the dependence of morphology and luminosity on environment . because galaxy properties depend not only on luminosity but also on morphology , it is clear that galaxy properties can not be determined solely by dark halo mass .
when morphology and luminosity are fixed , other physical properties , such as color , color - gradient , concentration , size , velocity dispersion , and star formation rate , are nearly independent of local density , without any break or feature .
the only feature is the sharp decrease of the late type fraction above the critical luminosity of about @xmath1 in the morphology versus luminosity relation .
weak residual dependences on environment include that of the color of late types ( bluer at lower density ) and of the @xmath2-@xmath3 relation of early types ( larger dispersion at higher density for bright galaxies ) .
the fraction of galaxies with early morphological type is a monotonically increasing function of local density and luminosity .
the morphology - density - luminosity relation , as measured in this work , should be a key constraint on galaxy formation models .
we demonstrate that the dependence on environment of the morphology of galaxies originates from variations in density on effective gaussian smoothing scales much smaller than @xmath4mpc .
we find that galaxy morphology varies both with density measured on an effective gaussian smoothing scale of @xmath5mpc and with distance to the nearest bright galaxy , particularly when the distance is about 0.2 @xmath6 mpc .
we propose as a mechanism that the morphology of galaxies in galaxy systems is transformed by the tidal force . |
high energy neutrino detection is one of the experimental challenges in particle astrophysics for the forthcoming years because it opens a new window to the regions of the universe that are otherwise shielded from us by large amounts of matter .
it is widely believed that one of the most appropriate techniques for neutrino detection consists of detecting the erenkov light from muons or showers produced by the neutrino interactions in underground water or ice .
this allows the instrumentation of large enough volumes to compensate for both the low neutrino cross section and the low fluxes expected .
there are several projects under way to build sufficiently large detectors to measure the expected signals from a variety of sources@xcite .
on the other hand many years ago it was suggested that deeply penetrating high energy particles , such as muons and neutrinos , initiate large horizontal air showers that can be detected at ground level @xcite . at large
zenith angles the electromagnetic part of ordinary air showers , initiated by cosmic rays of hadronic ( or electromagnetic ) nature , is attenuated by the atmosphere well before reaching ground level .
this should allow the identification of the showers initiated by these deeply penetrating particles .
recently there has been a proposal to build two 3000 km@xmath3 extensive air shower arrays , one in each hemisphere , to detect cosmic rays with energy above @xmath0ev ( the pierre auger project)@xcite .
each array consists of a hexagonal grid of water tanks , 3.5 m diameter and 1.2 m height , combined with an optical fluorescence detector .
the tanks are separated 1.5 km from each other and instrumented with photodetectors to detect the erenkov light emitted as photons and charged particles from the shower front cross it .
we will show that these detectors can also play an important and complementary role for detecting neutrinos of energy around @xmath4 ev and above .
neutrinos of these energies can not penetrate the earth so any detector searching them must look for events with zenith angles between near horizontal and vertical downgoing . for showers of sufficient energy
the array efficiency is high and the low target density for neutrinos provided by the atmosphere is compensated by the huge surface area of the planned array .
the observatory will complement other neutrino detectors in construction or planning . in this article
we show that the pierre auger observatories have an enormous potential to detect neutrinos in the eev range . in section
ii we firstly present some of the possible eev neutrino fluxes that have been discussed in the literature which will be used later in the evaluation of the event rates expected by the auger particle arrays . in section iii
we discuss the acceptance for the horizontal air showers .
firstly we obtain a rough but intuitive analytical expression for the acceptance of a large array , which we discuss in the context of the pierre auger project . in section
iii.a we present an acceptance calculation for the auger project based on a geometrical approach . in section iii.b an alternative and conservative calculation based on monte carlo simulation is presented . in section
iv we estimate the horizontal shower rates expected for the neutrino fluxes addressed in section ii , and section v is reserved for the conclusions .
we do not attempt to discuss event reconstruction or backgrounds which will be addressed elsewhere .
neutrinos of eev energies are likely to be directly produced together with cosmic rays of the highest energies whose origin is still a matter of speculation .
several possible sources of neutrinos in the eev region have been discussed in this context .
moreover , provided that the highest energy cosmic rays are extragalactic in origin , as it is currently believed , eev neutrinos have to be produced in their interactions with cosmic microwave photons .
we will use some of the neutrino flux predictions reported in the literature as reference calculations to evaluate the event rates that could be expected in each auger observatory . in this section
we briefly motivate these fluxes which are shown in fig .
1 compared with the expected atmospheric neutrino flux . active galactic nuclei ( agn ) , the most energetic objects known , emit most of their luminosity in gamma rays @xcite and may be the source of the highest energy cosmic rays @xcite .
multiwavelength observation of these sources leads us to believe they are powered by large accreting black holes , where shock fronts provide particle acceleration .
if , besides electrons , protons are also accelerated , as some models suggest , photoproduction of pions with the ambient light plays a crucial role ( see for example ref . @xcite and references therein ) .
the energetic gamma rays observed come from the decays of @xmath5 s and high energy neutrinos from the decays of charged pions become a signature of these models .
first models for neutrino production in agn assumed shock acceleration in the agn cores and predicted relatively flat fluxes up to energies of about @xmath6ev . for our event rate calculation
we select the prediction of ref .
@xcite , labelled agn-92c , which is quite similar to that in ref .
@xcite ( agn-95c ) .
there is however recent evidence that the gev to tev gamma ray emission observed from agn corresponds to the blazar class @xcite . in a unified agn description ,
the blazar class is identified as agn with jets of ultrarelativistic particles streaming out of their cores and pointing towards us .
most recent models for the proton blazars site the acceleration in the jets themselves .
photoproduced neutrinos are lorentz boosted to energies well above those predicted in the models of acceleration in the agn cores .
we use the prediction of ref .
@xcite ( labelled agn-95j ) which illustrates that the emitted neutrinos may extend well into the eev region in agreement with ref . @xcite .
we also consider the more speculative and uncertain models where the highest energy cosmic rays are produced by the decays of topological defects .
these objects emit massive x particles , predicted in grand unified theories , which decay and fragment into standard model particles .
there are a fair number of neutrino flux predictions from these models depending on the different time evolution of the effective injection rate of x particles per unit volume ( usually denoted with a parameter @xmath7 ) , the masses of the x particles themselves and other unknown parameters .
some models have already been constrained by bounds on horizontal showers induced by neutrinos @xcite and also by measurements of the 100 mev diffuse gamma ray background @xcite .
we select two fluxes illustrating the range of predictions in such models .
we use the prediction with @xmath8 in ref .
@xcite ( labelled td-92 ) and that given in ref .
@xcite which is labelled as td-96 .
td-92 is the lowest neutrino prediction given in ref .
@xcite and is not severely affected by experimental constraints .
lastly we consider two predictions for the high energy neutrinos produced in the interactions of cosmic rays with the cosmic microwave background : cmb-91 @xcite and cmb-93 corresponding to model 4 in ref .
they illustrate the variations that can be associated to various uncertainties intrinsic to these calculations .
these neutrinos are a direct result of the greisen - zatsepin - kuzmin cutoff and must be produced if the highest energy cosmic rays are of extragalactic origin .
unfortunately it is possible that the flux levels are too low to be observed either with the pierre auger observatory or with other neutrino detectors in planning or construction .
when the electron neutrino flux prediction is not explicitly available in any of the models used , we approximate it to be a factor of two below the muon neutrino flux as expected by naive channel counting in pion production and decay .
neutrinos produce showers in most interactions with the atmosphere which are of different nature depending on the process in consideration .
we consider both deep inelastic charged and neutral current interactions which always produce hadronic type showers . in the case of charged current electron neutrino interactions
the emerging electron contributes in addition a pure electromagnetic shower carrying a large fraction of the incoming particle energy .
we will ignore the resonant cross section because it is only significant near the peak of the cross section which occurs at an incoming neutrino energy of 6.4 pev , well below the region of high efficiency for the pierre auger project .
for a neutrino flux @xmath9 interacting through a process with differential cross section @xmath10 , where @xmath11 is the fraction of the incident particle energy transferred to the target , the event rate for horizontal showers can be obtained by a simple convolution : @xmath12 = n_a \rho_{air } \int_{e_{th}}^{\infty } de_{sh } \int_{0}^{1 } dy \ , \frac{d \phi_{\nu } } { de_{\nu } } ( e_{\nu } ) \ , \frac{d \sigma}{dy } ( e_{\nu},y ) \ , { \cal a } ( y , e_{\nu } ) \ ] ] where @xmath13 is avogadro s number and @xmath14 is the air density .
the energy integral corresponds to the shower energy @xmath15 which is related to the primary neutrino energy @xmath16 in a different way depending on the interaction being considered .
@xmath17 is a geometric acceptance , a function of shower energy , which corresponds to the volume and solid angle integrals for different shower positions and orientations with respect to the array .
the function is different for showers induced by charged current electron neutrino interactions from those arising in neutral current or muon neutrino interactions .
this is because hadronic and electromagnetic showers have differences in the particle distributions functions , particularly for muons . for showers of sufficient energy ,
incident with zenith angle @xmath18 , the effective volume for horizontal shower detection is given by @xmath19 , where @xmath20 is the surface area covered by the array , multiplied by the range of allowed positions for the first interaction point along the incident direction .
we can estimate this range approximating shower maximum as a disk of some effective radius @xmath21
. this radius should approximately be the maximum distance to the shower axis at which the particle density is high enough to trigger the detector tanks ( see below ) .
as the first interaction point is moved along the incident direction the intersection of this disk with the detector plane is an ellipse with major axis @xmath22 with @xmath23 .
the projection of this axis onto shower direction ( @xmath24 ) is the wanted range ( see fig .
we estimate @xmath17 integrating this volume over the possible solid angle orientations of the shower , @xmath25 , and restrict the integration to horizontal showers i.e. @xmath26 .
approximating @xmath27 we obtain a simple analytic expression in terms of @xmath21 .
@xmath28 for @xmath29 the ellipse axis @xmath30 exceeds @xmath31 , the average size of the array the diameter of the pierre auger array is approximately @xmath32 and @xmath33 . ] .
the last term in the brackets of eq .
( 2 ) represents a small correction obtained when restricting the range of the interaction point to be below the longitudinal size of the array .
most importantly the acceptance is seen to depend on the array surface area and the shower radius .
the estimate makes the assumption that showers with shower maximum intercepting the array are detected with close to @xmath34 efficiency .
the effective radius has to be chosen to match this requirement which will only hold for energies above a given threshold .
it is conservative in the sense that it ignores the fact that showers can trigger the detector without having shower maximum intercepting the array , or even without the shower axis going through the array ( i.e. completely horizontal showers ) .
it is easy to see that the acceptance for the auger particle arrays is comparable to other neutrino detectors in planning @xcite .
since design requirements have led to a tank size that allows near @xmath35 efficiency for vertical showers of energy above @xmath4 ev @xcite , the expected signal in the tanks should be large enough to trigger up to distances from shower axis of order the separation between the tanks , @xmath36 km .
it is thus conservative to expect similar efficiencies for horizontal showers above @xmath4 ev as they should not differ that much from vertical ones at detector level and to approximate @xmath37 km .
the water erenkov technique is particularly well suited for horizontal showers .
the transverse separation between detectors will be substantially reduced for near horizontal showers and the extra depth of the tanks in the horizontal mode should enhance the signal from the muons in the shower , compensating the reduction in transverse area of the tank to the incident direction .
taking @xmath38 km ( @xmath39 ) we obtain an estimate for the auger array acceptance of showers of energy above @xmath40 ev , @xmath41 which when multiplied by an air density @xmath42 gives @xmath43 .
underground high energy neutrino detectors in planning aim towards an active volume in the range of @xmath44 of water @xcite . in models where most of the events are due to neutrinos well above the pev region ,
the earth will be opaque to them .
the corresponding acceptance for contained events is of order @xmath45 , illustrating how the pierre auger project may come into play .
the above estimate of the acceptance does not display the energy dependence but it illustrates how it depends on the effective shower radius , which in turn must depend on energy . as it happens for vertical showers , the acceptance must show a kind of threshold behavior , since sufficiently low energy showers can not be detected with such a sparse array .
the dependence of the acceptance on shower energy has been introduced in two alternative ways . in a more geometrical approach
we have followed on the above line of thinking considering in more detail the shower depth development and the lateral distribution functions for electrons and positrons . as an alternative we have simulated showers at different positions with respect to the detector in a monte carlo approach .
the calculation treats electromagnetic and hadronic cascades independently using the conventional nkg charged particle lateral distribution functions @xcite normalized with the gaisser @xcite ( greisen @xcite ) parametrization for the total number of electrons and positrons in hadronic ( electromagnetic ) cascades .
we assume that triggering requirements for each tank can be specified as fixed numbers for the electron density in the electromagnetic and hadronic showers : @xmath46 .
the relevant quantity for the tank signal is the erenkov light , proportional to the charged particle tracklength , which is in turn approximately proportional to energy deposition .
muons travel through the tank depositing about @xmath47 mev/(g @xmath48 ) while most electrons and photons are stopped in the tanks , depositing all their energy .
we use @xmath46 because there are convenient parameterizations in the literature for this quantity .
the relation between @xmath46 and energy deposition is however dependent on the shower position and can only be done in an approximate manner .
the above assumption naturally defines the active region of a shower , the volume within which the electron density exceeds the given threshold @xmath46 .
the region is bounded by a cigar shaped contour plot of the three dimensional electron distribution function at @xmath46 . for the solid angle - volume acceptance integration , we consider the number of tanks contained in the intersection of these active volumes with the detector plane as the shower directions and first interaction point are varied .
the electron - positron density at these tanks is by definition above threshold . at typical large distances
( of order a kilometer ) to the shower axis , we estimate that an electron density in the range 0.6 - 1.2 m@xmath49 ( 0.3 - 0.5 m@xmath49 ) for horizontal electromagnetic ( hadronic ) showers is equivalent to an energy deposition of 500 mev in a tank ( corresponding to about two vertical muons ) .
this is considered sufficient for shower detection @xcite . approximating the intersections of these active regions as rectangles allows a complex but analytical solution to the problem if the only orientations and positions of the rectangles considered in the integration are those which contain at least @xmath50 tanks in the same row ( see fig .
3 ) . as a result
we obtain the acceptance curves as a function of shower energy for each shower distribution function which only depend on the parameter @xmath46 .
the important advantage of this calculational method against the obvious alternative which involves monte carlo simulation ( addressed below ) is computing speed .
the approach converts in straightforward the otherwise lengthy evaluation of the effect of changing the input parameters in the calculation such as trigger conditions , tank thresholds , parameterizations of shower distribution functions and even array spacing . in fig .
4 we show the acceptance results for a trigger of at least @xmath51 aligned tanks , having the conservative electron density of 1.2 m@xmath49 ( 0.5 m@xmath49 ) for electromagnetic ( hadronic ) showers and considering only showers with zenith angle higher than 75 degrees . for each shower type
there are two curves , one considers showers with axis falling within the array and the second also includes an approximate calculation of the triggering showers with axis not going through the array .
it is important to note that these showers make most of the contribution to the acceptance in the high energy limit ( corresponding to huge showers flying parallel and on top of the array ) . besides using @xmath46 as the parameter for tank threshold ,
the geometrical algorithm described above has a potentially more important drawback . by using parameterizations for air showers
, it ignores the effect of the ground on shower development . the simplification can be thought to grossly overestimate the acceptance .
using egs4 with constant density air , we have simulated the effect of the ground for very inclined electromagnetic showers of energies up to @xmath52 . in spite of the reduction in the lateral distribution observed
after the core hits the ground , there is only a very small reduction in the acceptance assuming the results can be extrapolated to higher energies . in any case
we have decided to contrast the acceptance results against a totally different calculational approach , which we chose to remedy these drawbacks in a conservative manner . that way we hope to bracket a prediction for the horizontal shower acceptance of the auger project .
we have performed a totally independent and conservative calculation with a hybrid monte carlo and parametrization method .
the approach uses the simulated curves of energy deposition versus distance from shower axis in the pierre auger water tanks for a @xmath0ev shower at different depths .
these results are then simply scaled with shower energy .
we have restricted the range of depths for the shower to the interval [ 500,1600 ] g @xmath48 and we only consider showers with the shower axis falling within the array . in our effort to be conservative
we completely neglect the contributions from particles after the shower core hits the ground .
we have then generated horizontal events by allowing the starting point and direction ( @xmath53 ) of the shower to vary in a random way , only restricted by the array size and a maximum height of 3 km . in the simulation of the acceptance
, events are detected when the shower deposits more than 500 mev of energy in each of four or more tanks .
we have used an average parameterization assuming half of the energy induces a hadronic shower and the other half an electromagnetic shower . besides correcting drawbacks in the geometrical approach , the method has the advantage of allowing the study of the different patterns of hit tanks on an event by event basis@xcite .
the results are shown as points in fig . 4 for some discrete energies .
given the fundamental differences between the two approaches , it is not possible to test one against the other .
nevertheless it is fair to remark that the agreement is rather good and that the differences of results can be interpreted on the basis of the different inputs intrinsic to each method .
we thus take the monte carlo approach as a lower bound of the acceptance and the difference between the two approaches can be considered as an indication of the degree of uncertainty . for the detection of the horizontal showers included in the acceptance calculations described here
, only directional reconstruction and an ability to separate the electromagnetic part of the shower from the muonic component should be sufficient in principle .
nevertheless if the long range of the muons produced in the horizontal showers of hadronic nature is also taken into account , it can be argued that the acceptance may exceed that actually calculated in this work .
the muon component effectively enhances the length of the active region of the shower , what must increase the acceptance for deeply penetrating particles .
the identification of such showers would become however more complicated , because they look more like those induced by cosmic rays at very large zenith angles .
to calculate event rates we use the deep inelastic charged and neutral current cross - section from two sets of structure functions mrs(r1 ) and grv ( @xmath54 renormalization scheme ) parton distributions @xcite . with increasing energy , neutrinos interact with partons carrying a lower fraction @xmath55 of nucleon momentum which extends beyond the kinematical limits of the parton density parameterizations .
for the first set we extrapolate to low @xmath55 beyond validity of the parameterization , using the slope of @xmath56 , at the lowest @xmath55 permitted ( in this case @xmath57 ) , where @xmath58 is the standard parton distribution .
this is consistent with extrapolations based on the leading log approximation @xcite .
the second set , grv , can be cautiously used on its own for low @xmath55 .
table 1 displays the event rates for different neutrinos fluxes , calculated with eq .
( 1 ) for the each of the two cross sections .
the first entry corresponds to the result of our conservative approach based on the simulation and the higher one to the acceptance results calculated in the geometrical approach including showers with axis not intercepting the array .
recent models of proton acceleration in agn jets ( agn-95j ) and some models of decays of topological defects ( such as td-92 ) predict neutrino fluxes giving measurable rates in the most conservative assumptions . for neutrinos from agn cores
the majority of the detected showers lie in the @xmath59 ev region .
this corresponds to the threshold region for the acceptance curves where the efficiency is low and are very sensitive to trigger conditions . for these neutrino fluxes ,
a @xmath44 conventional underground neutrino detector is expected to give more events .
the detection of neutrinos from interactions of cosmic rays with the cosmic ray background is unfortunately only marginal for the most optimistic predictions .
it should be mentioned that conventional muon underground detectors in planning will have similar difficulties , if not higher , in the detection of these neutrinos .
the pierre auger project can be made sensitive to ultra high energy neutrino fluxes through horizontal showers if an appropriate trigger is implemented .
its acceptance for detecting contained neutrino events of energy above @xmath60 can be very large and comparable to neutrino telescopes in planning .
the highest efficiency for horizontal shower detection with the pierre auger project is expected at energies about @xmath61 , what makes it in principle a tool for the search of the neutrinos from interactions of cosmic rays with the cosmic microwave background .
the auger array can detect the very high energy neutrinos from the decays of topological defects in some predictions . in any case the plethora of topological defect models will be further constrained by the neutrino detection capabilities of the auger observatories .
most recent predictions for neutrinos produced by protons accelerated in the jets of agn have also prospects to be detected as horizontal shower events in the auger particle arrays .
we thank j. alvarez muiz , j.j .
blanco pillado , p. billoir , f. halzen , a. letessier - selvon , k. mannheim , j. mathews , d. nitz , r. protheroe , c. pryke , g. sigl , r. vzquez and a.a .
watson for a number of discussions and suggestions .
this work was supported in part by cicyt ( aen96 - 1673 ) and by xunta de galicia ( xuga-20604a96 ) .
k.s.c . was supported by nsf grant phy 9506751 `` research experiences for undergraduates '' .
gaisser , f. halzen and t. stanev , phys . rep .
238 ( 1995 ) 173 .
berezinsky and a.yu .
smirnov , _
astrophys.space science * 32 * ( 1975 ) 461_. design report of the pierre auger collaboration , fermilab report , february 1997 .
thompson et al . , _ ap . j. suppl . _
* 101 * ( 1995 ) 259 ; _ ap . j. suppl .
_ in press ( 1996 ) .
p. biermann and p.l .
stritmatter , _ ap . j. _ * 322 * ( 1987 ) 643 ; f.w .
stecker and m.h .
salamon , preprint astro - ph/9501064 , to appear in space sci .
a.p.szabo , and r.j .
protheroe , in proc .
high energy neutrino astrophysics workshop ( u. hawaii , march 1992 ) , ed .
stenger , j.l . learned , s.pakvasa and x. tata , world scientific , 1993 , p. 24 .
mattox et al . , to appear in _ ap .
j. _ * 481 * ( 1997 ) .
k. mannheim , astrop .
phys . 3 ( 1995 ) 295 .
protheroe , in proc .
iau colloq .
163 , accretion phenomena and related outflows , ed .
d. wickramasinghe et al .
, in press ( 1996 ) .
blanco - pillado , r.a .
vzquez and e. zas , _ phys .
* 78 * ( 1997 ) 3614 .
protheroe and t. stanev , _ phys .
* 77 * ( 1996 ) 3708 and _ erratum _ * 78 * ( 1997 ) 3420 .
p. battacharjee , c.t .
hill , d.n .
schramm , phys .
* 69 * ( 1992 ) 567 .
g. sigl et al . ,
astro - ph/9605158 submitted to ap .
j. lett .
stecker , c. done , m.h .
salamon and p. sommers , phys .
* 66 * ( 1991 ) 2697 .
s. yoshida and m. teshima , prog . of theo .
( 1993 ) 833 .
g. parente and e. zas , proc . of the high energy neutrino workshop , pp .
435 , venice , february 1996 , astro - ph/9606091
. f. halzen , proc . of the high energy neutrino workshop , pp .
499 , venice february 1996 .
k. greisen , prog .
in cosmic ray phys . , ed . j.g .
wilson , vol.iii , p.1 , north holland publ .
co. ( 1956 ) ; k. kamata and j. nishimura , prog .
( kyoto ) suppl .
* 6 * ( 1958 ) 93 .
gaisser , cosmic rays and particle physics , cambridge univ . press , 1990 and references therein .
capelle and j.w .
cronin , 1996 ( unpublished ) .
martin , r.g .
roberts and w.j .
stirling , _ phys .
( 1996 ) 419 _ ; m. glck , e. reya and a. vogt , _ z phys .
cparticles and fields * 67 * ( 1995 ) 433_. g. parente and e. zas .
25th intern .
cosmic ray conf .
potgieter et al .
durban 1997 .
vol.7 p. 109 .
figure 2 : schematic picture of a horizontal shower , illustrating the disk of radius @xmath21 at shower maximum and the cylinder it spans as the first interaction point is shifted along the incident direction .
the intersection of the cylinder is an ellipse of major axis @xmath30 ( see text ) .
figure 3 : illustration of the intersection of the `` active part '' of the shower ( as described in the text ) and the detector plane used in the geometrical approach .
the intersection , which is close to an ellipse , is approximated by a rectangle as illustrated .
the dots correspond to the detector tanks .
figure 4 : acceptance of the pierre auger detector to near horizontal showers ( @xmath63 degrees ) .
volume units are @xmath64 of water equivalent .
crosses are for the results in the monte carlo approach , lines are correspond to the geometrical integration for electromagnetic showers ( dashed ) and hadronic ( solid ) . the lower set of curves corresponds to only showers with axis falling in the array and the upper set takes all showers into consideration . | we show that the auger air shower array has the potential to detect neutrinos of energies in the @xmath0ev range through horizontal air showers . assuming some simple conservative trigger requirements we obtain the acceptance for horizontal air showers as induced by high energy neutrinos by two alternative methods and we then give the expected event rates for a variety of neutrino fluxes as predicted in different models which are used for reference .
# 1#2#3 _ ann .
phys .
( ny ) _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ astrophys .
j. _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ astrophys .
j. lett .
_ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ acta .
phys . pol .
_ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ ann .
rev .
nucl . part .
sci . _ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ computer phys .
comm . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ erratum _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ ibid . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ j. math .
phys . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ int
. j. mod .
phys .
_ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ jetp lett .
_ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ j. phys .
g. _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ mod .
phys .
lett . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ nature ( london ) _ * # 1 * ( # 2 ) # 3 . #
1#2#3 _ nuovo cim .
_ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ nucl .
instr .
meth . _
* # 1 * ( # 2 ) # 3 .
# 1#2#3 _ nucl .
phys . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ proc . cam .
phil .
soc . _
* # 1 * ( # 2 ) # 3 .
# 1#2#3 _ phys .
lett . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ phys .
rep . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ phys .
rev . _ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ phys .
rev .
lett . _ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ proc .
roy .
soc . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ prog .
th .
phys . _ * # 1 * ( # 2 ) # 3 .
ps .
# 1#2#3 _ physica scripta _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ rev .
mod .
phys . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ rep .
prog .
phys . _ * # 1 * ( # 2 ) # 3 .
# 1#2#3 _ sov .
j. nucl .
phys . _ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ sov .
. jept _ * # 1 * ( # 2 ) # 3 . # 1#2#3 _ sov .
phys .-
usp . _
* # 1 * ( # 2 ) # 3 .
# 1#2#3 _ zeit .
phys .
_ * # 1 * ( # 2 ) # 3 .
* on the detection of ultra high energy neutrinos with the auger observatory * + k. s. capelle@xmath1 , j. w. cronin@xmath1 , g. parente@xmath2 and e. zas@xmath2 @xmath1__department of physics and enrico fermi institute , _
_ _ university of chicago , chicago , il 60637 , usa _ @xmath2__departamento de fsica de partculas , _ _ _ universidade de santiago de compostela , e-15706 santiago , spain _ pacs numbers : 95.85.ry , 96.40.tv , 96.40.pq , 98.70.sa |
in this summer school , we were given many examples of glassy systems , glassy dynamics , and glass transitions , even though a proper definition of the word ` glassy ' was not really provided .
however , all glassy materials share the property that their relaxation times are extremely large compared to the time scale of a typical experiment , at least in a part of their phase diagram . for practical purposes , they are thus out of equilibrium , meaning that in principle the whole sample history is relevant to a description of their physical properties .
this paper is dedicated to the study of some specific histories applied to various glassy materials . as physicists , we want to study the simplest histories that allow for an understanding of all the relevant mechanisms at work .
hopefully , an understanding of simple protocols will also allow for prediction or calculation of the behavior resulting from increasingly elaborate procedures . over the last decades
, the experiment most often performed has been the _ simple aging _ experiment , see fig .
[ phasediag ] .
the system is quenched from a non - glassy part of the phase diagram , @xmath0 , into the glassy phase , @xmath1 .
the system relaxation time is so large that all its physical properties continue to evolve slowly with time ( aging ) .
this phenomenon has been known for a long time in the field of structural glasses @xcite before its rediscovery in the field of spin glasses @xcite .
interestingly , aging is observed in a still increasing number of experimental systems such as soft materials ( like pastes @xcite , colloidal suspensions @xcite , or clays @xcite ) , dipolar glasses @xcite , disordered ferromagnets @xcite and ferroelectrics @xcite , granular matter @xcite , superconductors @xcite , etc .
loss of stationarity is best illustrated by the study of two - time quantities .
one typically computes the correlation between times @xmath2 and @xmath3 , @xmath4 , or the response of the system at time @xmath2 to a perturbation applied at @xmath3 , or equivalently the time evolution of a susceptibility at frequency @xmath5 , @xmath6 . in most common experimental regimes ,
two - time quantities scale as @xmath7 , @xmath8 , where @xmath9 denotes a scaling function .
this scaling indicates that after a time @xmath2 the only relevant time scale in the system is the time @xmath2 itself . in order to probe the dynamics of the glassy phase in greater detail , more elaborate and systematic experimental protocols have been performed @xcite , in which some external parameters are shifted , or cycled , during the experiment , see fig .
[ phasediag ] .
such experiments reveal spectacular new phenomena , such as rejuvenation and memory effects . such new effects must be accounted for by any theory of aging , possibly allowing for discrimination between different theoretical approaches to glasses and aging phenomena . in addition , more detailed experiments may allow for discrimination between different families of glassy systems , and thus may help theoreticians refine their description of specific glassy systems .
a large number of recent experimental papers are dedicated to such experiments , on a wide variety of systems , making the subject a very active one . by contrast , these protocols are barely mentioned in the classic theoretical review in the field @xcite and it is one of this paper s purpose to fill this little gap . in the first part of the paper we review briefly the basic experimental facts .
next we discuss two mean - field theoretic approaches to the problem .
we then formulate the rudiments of a simple , but fairly robust , phenomenology in terms of length scales which grow with time and discuss realizations of this scenario .
last , we show that a nice account of both recent experiments performed on an anderson insulator and on a colloidal suspension can be given in terms of such growing length scales .
in this section , we present the main experimental facts observed when the protocols of fig .
[ phasediag ] are actually performed .
this will allow us also to define precisely the vocabulary used throughout the paper .
we use our own data to describe these phenomena , but emphasize that a similar phenomenology has been observed in many different systems .
such experiments were first performed on spin glasses ( rejuvenation and memory effects @xcite ) and polymer glasses ( mechanical rejuvenation @xcite and kovacs effect @xcite ) in temperature shift or cycling experiments .
thus , we adopt ` temperature ' as a control parameter , but also discuss the case of other control parameters .
the degree to which different such external parameters are equivalent is a completely open question .
it was recently asked in the present context @xcite , as a part of a more general research line @xcite . here , we take a pragmatic approach and elaborate on experimental similarities . sections
[ rejuvenation ] , [ underaging ] , and [ kovacs ] deal with effects encountered in shift experiments , while section [ memory ] deals with cycling experiments .
first consider a shift experiment , see fig .
[ phasediag ] .
the system is quenched at initial time @xmath10 from a high temperature to a temperature @xmath11 in the glassy phase .
for @xmath12 , the temperature is kept constant . up to this point , this is a simple aging experiment , manifested by the slow evolution of physical quantities . such typical slow evolution is shown in the left part of fig .
[ effects1 ] where a quantity analogous to a magnetic susceptibility at given frequency is computed in the numerical simulation of a microscopic spin glass model @xcite . at @xmath13 ,
the temperature is shifted to @xmath14 . as seen in fig .
[ effects1 ] , aging is restarted by a negative shift ( @xmath15 ) in the sense that the resulting curve is similar to that obtained in a direct quench to @xmath14 . this restart of the dynamics is called _ rejuvenation effect _ because the time @xmath16 spent at @xmath11the sample ` age'seems to have no influence on the dynamics following the shift .
the same effect is obtained if @xmath17 .
we note also that the term ` rejuvenation ' was first employed to describe the effect of large stresses on the aging of polymers @xcite . [ cols="^,^ " , ] obviously , if @xmath18 , no rejuvenation takes place , meaning that in order for rejuvenation to be observed in a shift experiment , @xmath19 must be ` large enough ' . for small or intermediate @xmath19 , a phenomenon recently called _ overaging _ is observed @xcite , when @xmath20 .
this is illustrated in fig .
[ effects2 ] ( left ) , where results obtained with a polymer are presented @xcite . in this figure ,
two - time linear response to a stress step ( creep compliance ) after a shift from @xmath21 to @xmath22 is compared to that obtained in a simple aging experiment at @xmath23 with the same aging time .
the response after the shift is slower than that obtained in the the simple aging experiment , and the system looks ` older ' , or ` overaged ' .
similarly , an _ underaging _ would be obtained if @xmath24 .
these effects were also observed experimentally in a colloidal suspension submitted to a transient oscillatory shear @xcite , as well as in temperature shift protocols in experimental @xcite and numerical @xcite studies of spin glasses . a _ memory effect _ takes place in the shift protocol when @xmath17 .
it was first observed by kovacs in polymers @xcite . to distinguish it from a second memory effect ( see below ) , this effect was called _
kovacs effect _ in ref .
kovacs measured the specific volume , @xmath25 , of the polymer during the experiment , but other physical quantities ( index of refraction , energy density , ... ) can be investigated .
the effect is particularly striking when @xmath16 is chosen so that @xmath26 , which means that immediately after the quench , the volume has already reached its equilibrium value at the new temperature .
hence , a naive expectation would be that @xmath27 .
instead kovacs observed a nonmonotonic variation of the volume showing that the system has some _ memory _ of its state at the initial temperature . this experiment is reproduced in fig .
[ effects2 ] using the energy density of a microscopic spin glass model in a numerical simulation @xcite .
similar results have been obtained in supercooled liquids @xcite , granular materials @xcite , foams @xcite or dipolar glasses @xcite .
let us describe the continuation of the experiment shown in fig .
[ effects1 ] . at time @xmath28
, @xmath29 is shifted backed to its initial value @xmath11 .
after a very short transient , @xmath6 resumes its evolution as if there had been no aging at @xmath14 .
the system has a _ memory _ of the first stage , despite the strong rejuvenation observed in the intermediate stage of the cycle .
simultaneous observation of rejuvenation and memory is spectacularly demonstrated in the ` dip experiment ' @xcite .
the protocol is essentially a cycling experiment in which the temperature is decreased at a fixed , finite rate ( instead of at an infinite rate as in an idealized cycle ) .
the ramp from high temperature stops at @xmath11 , where the temperature is kept constant a time @xmath16 after which cooling is resumed . in this context rejuvenation
means that the further evolution of the system is almost the same with or without the stop at @xmath11 .
the temperature is then raised back at a constant rate .
memory means that near @xmath11 , the system ` remembers ' its stop @xcite and behaves differently from a system not held at @xmath11 .
however , this experiment is less simple to analyze theoretically because it mixes cooling rate effects with rejuvenation and memory .
for this reason , we stick to cycles in what follows .
this quick experimental review shows that a number of effects are both highly non - trivial as well as generically observed in a wide range of materials , suggesting that these phenomena are intrinsic to nonequilibrium glassy dynamics .
this has two immediate consequences .
( 1 ) any phenomenological theory of aging must account for these effects in addition to the results of simple aging experiments .
( 2 ) the phenomenology should be based upon general considerations , which themselves proscribe the range of experimental systems to which the theory applies .
ideally this would be the large range of quite different experimental systems in which glassy dynamics is observed .
as reviewed in ref .
@xcite , theoretical approaches to aging phenomena can be classified in three large families .
we analyze two mean - field ones in this subsection in view of the above experimental facts . a first approach , made popular through the ` trap model ' formulated in ref .
@xcite , considers the dynamics from the point of view of phase space and describes the dynamics of a point particle evolving through a given energy landscape .
the reduction of a many - body system to a one - particle problem classifies this approach in the mean - field family .
the ` complexity ' stems from an assumed large distribution for the free energy of the metastable states , typically chosen to be exponential , @xmath30 . if one further assumes that energy barriers are related to the free energy of the states , one gets the distribution of trapping times @xmath31 , the first moment of which diverges for @xmath32 .
the absence of a mean trapping time then results in the typical @xmath33 scaling of two - time quantities .
obtaining this scaling in a one - body problem is a remarkable result .
the crucial drawback of the approach is , obviously , the lack of a precise interpretation of what ` traps ' are . to account for the effects described above , the model was phenomenologically extended to multi - trap models @xcite , each level having its own glass temperature @xmath34 .
the phenomenology of cycles is straightforward .
when the temperature is lowered , new levels of the hierarchy start to age ( rejuvenation effect ) , while higher levels are completely frozen and unaffected by the stay at low temperature ( memory ) .
this ` hierarchical phase space picture ' has often been invoked by the saclay group to interpret experiments on spin glasses @xcite .
furthermore , numerical simulations of concrete realizations of multi - trap models confirm that this picture satisfactorily reproduces the experiments @xcite . as a similar realization of this picture ,
the sinai model was recently studied in this context @xcite .
a more microscopic approach can also be used to describe aging @xcite .
the aim is to solve exactly the dynamics starting from the hamiltonian of a glassy system .
this ambitious program has been successful for systems with infinite - range interactions .
they are thus mean - field realizations of realistic systems .
simple aging experiments can be accounted for , as described in detail in cugliandolo s lectures @xcite .
these solvable models have proved to be extremely rich , but by construction they are incomplete descriptions since real space is completely ignored .
two families of models have emerged .
schematically , the first has the phenomenology of structural glasses
an example is the @xmath35-spin model and typically exhibits @xmath33 scalings @xcite .
the second family is closer to spin glasses the prototype being the sherrington - kirkpatrick model .
the behavior of the latter in simple aging experiments is quite involved , since it displays ` dynamic ultrametricity ' @xcite . without giving the details @xcite
, this implies a complex scaling of two - time functions , with the presence of a continuous hierarchy of diverging time scales .
when submitted to shifts and cycles , this second family ( but not the first one ) was shown to exhibit rejuvenation and memory asymptotically @xcite .
it would be useful to have simulations of these models in order to go beyond the asymptotic analysis of ref .
@xcite and to get some comparison to experiment . within this approach , the glassy effects of the previous section
are explained through the existence of a hierarchy of time scales . however , as honestly noted in the conclusion of ref .
@xcite , a strong drawback is that the dynamic ultrametricity on which the whole interpretation relies is incompatible with experiments .
obviously this weakens the general validity of these results .
a third family of models describing slow dynamics focuses directly on spatial aspects @xcite .
simple aging experiments are explained with reference to a characteristic length scale , @xmath36 , which grows with time .
the physical content of @xmath36 is that on scales smaller than @xmath37 the system appears equilibrated while on larger scales it does not .
a pure ising ferromagnet quenched from its paramagnetic to its ferromagnetic phase provides a simple example of coarsening dynamics .
in the low temperature phase , the ferromagnet has two equilibrium states , with magnetizations @xmath38 and @xmath39 .
as time passes , spatial domains of these equilibrium phases develop and coarsen , with a typical associated length scale @xmath36 .
this domain growth is driven by surface tension of domain walls .
the dynamic scaling hypothesis is that the only length scale involved in this process is the domain size @xmath36 itself . in a pure ferromagnet ,
the growth law is temperature independent @xcite , @xmath40 , and this leads to @xmath33 scaling of two - time correlators
. this phenomenology does not provide for the rejuvenation and memory effects discussed above and indeed such effects are not observed in ferromagnets . for instance , consider rejuvenation .
thermal reequilibration within the domains is quasi - instantaneous since thermal fluctuations in the ferromagnetic phase are short - range , of typical size @xmath41 .
also , since lowering temperature during domain growth leaves the growth rate unchanged , after a temperature cycle @xmath42 , there can not be regions initially equilibrated at @xmath11 that remain unchanged during the cycle .
however , coarsening in a disordered system can not be analogous to that in a pure system @xcite .
coarsened domains will no longer be simple convex objects with energy decreasing monotonically with curvature . this can be seen readily by considering a ferromagnet with some added impurities .
walls separating @xmath38 and @xmath39 domains will tend to avoid unusually strong bonds and will tend to be pinned at unusually weak bonds .
in such a disordered system , domain growth will be slower than it is in a pure ferromagnet . in the case of high disorder ,
domain growth is likely activated , leading to a characteristic domain size @xmath43 @xcite .
secondly , in important contrast to pure systems , renormalization of a disordered system s hamiltonian yields statistically similar hamiltonians . in particular
this means that we should not expect that the equilibrium states in the low temperature phase of a disordered system at temperatures @xmath11 and @xmath14 are related in any simple way .
a simple example is given by the effective interaction between two spins on the opposite vertices of a square . just as the sign of the effective interaction between these two spins can change with a perturbation in bond strength , so too can it change with a perturbation in temperature @xcite .
these differences stated , one can ask whether there is a phenomenological coarsening theory for aging in glassy materials analogous to that in systems like pure ferromagnets .
we formulate such a ` minimal ' phenomenology in terms of a growing length scale .
it must contain the following features . _
( i ) existence of a growing coherence length @xmath37_. this coherence length has the same physical content as described above : objects smaller than @xmath37 are quasi - equilibrated while larger objects still retain their nonequilibrium initial conditions . a more precise definition of the ` objects ' is not necessary at this point . for clarity we will draw compact domains in our cartoons
furthermore , the coherence length must grow slowly with time so that equilibrium domains of size @xmath37 remain fixed on time scales associated with objects smaller than @xmath37 .
this first assumption is very natural for some systems , e.g. , disordered ferromagnets . yet for other systems it is very non - trivial .
for instance , no typical length scale has been invoked yet to explain the aging of supercooled liquids or colloidal suspensions . _
( ii ) sensitivity of equilibrium on all length scales .
_ obviously , the specific equilibrium state of the system depends upon the particular values of the control parameters , but sensitivity is required in order to provide for the rejuvenation effects .
of course , even a simple two - level system has to readapt its boltzmann weights upon a temperature change @xcite .
there remains to be understood what ` levels ' are in a realistic system , and how those evolve with temperature .
in fact as discussed above , sensitivity of the equilibrium state to control parameter values is a natural consequence of a renormalization procedure in a disordered system .
note that there are nevertheless questions outstanding .
in particular , which if any glassy systems be described in terms of a glassy phase fixed point ? and even if some can be , is the sort of ` chaos ' that one gets automatically in fact responsible for the observed rejuvenation and memory phenomena in glassy materials ?
we will return to these points shortly . _
( iii ) separation of length scales .
_ the length at which a system is equilibrated after a given wait time depends strongly on external parameters .
memory effects are a consequence of such dependence .
this is very natural for the thermally activated domain growth expected in disordered systems @xcite : @xmath44 where @xmath2 is the time needed to equilibrate the length scale @xmath37 , @xmath45 a microscopic time scale , and @xmath46 an activation energy that in principle can depend on the coherence length itself and on temperature .
the dots stand for possible control parameters in addition to @xmath29 . if @xmath47 , the growth law is logarithmic and ` super - arrhenius ' : @xmath48 .
if instead barriers grow logarithmically , @xmath49 , one gets a @xmath29-dependent power law growth , @xmath50 . in either case , thermal activation ( eq .
( [ equation ] ) ) implies that the range of length scales ` active ' in a given time window depends strongly on @xmath29 and , more generally , on other external parameters in systems with glassy dynamics @xcite .
consider the cycling experiment @xmath51 of fig .
[ phasediag ] .
again we write the control parameter as a temperature @xmath29 .
the three above assumptions lead to the cartoon of fig .
[ cartoon ] which we now explain . for times
@xmath52 , this is a simple aging experiment at @xmath11 .
feature ( i ) entails that the system equilibrates up to a growing coherence length , @xmath53 , leading to a @xmath54 scaling of two - time functions .
this is the usual domain growth picture of aging .
small length scales , @xmath55 , with characteristic times @xmath56 have reached equilibrium at @xmath11 , while big ones , @xmath57 with characteristic times @xmath58 still retain their initial nonequilibrium state at @xmath59 , see fig .
[ cartoon ] .
then the temperature is shifted to @xmath60 at time @xmath16 with @xmath61 ` large ' . feature ( ii )
entails that all length scales , @xmath62 , must adapt to the new temperature @xmath14 .
this occurs through the growth of a new coherence length , @xmath63 , as shown by a dashed line in fig .
[ cartoon ] .
aging is thus restarted from scratch and _ rejuvenation due to the reequilibration of small length scales _ is observed . at time @xmath64 , the system is shifted back to @xmath1 .
due to ( iii ) , @xmath60 implies a much slower growth of @xmath37 during the intermediate stage .
hence , for @xmath65 , as is often the case in experiments , one has @xmath66 and the behavior for @xmath67 is due to three types of length scales .
small length scales , @xmath68 , have to reequilibrate at @xmath11 .
intermediate lengths , @xmath69 , are already equilibrated at @xmath11 . _
it is in these length scales that the memory resides_. large lengths , @xmath70 , still have to equilibrate at @xmath11 .
hence , after a short transient of duration @xmath71 given by @xmath72 , aging proceeds as the continuation of the first stage .
note that under the assumption of activated dynamics , simple inequality of length scales corresponds to a large degree of inequality in terms of time scales . if one chooses instead @xmath17 , the essential situation should be unchanged , provided that the condition @xmath66 is fulfilled , see fig .
[ cartoon ] .
this property is called ` symmetrical effect ' in the field of spin glasses @xcite .
note , however , that this condition on the lengths is extremely difficult to satisfy due to feature ( iii ) .
therefore , in most experiments with @xmath17 the equilibrated length grown at @xmath14 is larger than that grown at @xmath11 .
all memory is subsequently erased , resulting in an apparent asymmetry between positive and negative cycles .
let us finally turn to kovacs experiments .
two types of length scales must be distinguished in order to understand this effect .
small scales , @xmath73 , are equilibrated at @xmath11 but must adapt to a new , _ higher _ , temperature @xmath14 .
large lengths , @xmath70 , in contrast , are still in their nonequilibrium high temperature initial state and must adapt to a _
smaller _ temperature @xmath14 .
if the energy density @xmath74 is computed , as in fig . [ effects2 ] ( right ) , small length scales contribute to increase @xmath74 , while large length scales contribute to decrease @xmath74 .
this accounts in simple terms for the kovacs effect .
schematic discussions based on ( i ) , ( ii ) , and ( iii ) like that above allow for qualitative prediction of the outcome of given experimental protocols . however , while ( i ) , ( ii ) and ( iii ) are natural features for a phenomenological theory of coarsening in a glassy system , in the above they are put in by hand .
consequently we now discuss models that explicitly realize these properties .
the scaling , or droplet , model of spin glasses @xcite possesses these three features and provides quantitative , experimentally testable predictions . these predictions are clearest in the case of ising spin glasses , which we discuss here . at a given temperature in the spin glass phase ,
the assumptions made by the scaling picture resemble those made in the case of pure ferromagnetic systems , but differ due to the disorder in a spin glass .
the initial assumption of the droplet model is that ising spin glasses have two equilibrium states , related by spin flip symmetry .
spin glass dynamics are then described in terms of ` droplets ' , low lying excitations about these states @xcite .
because of disorder , the boundaries of low lying excitations wander in order to take advantage of bonds not satisfied in the ground state and to avoid satisfied bonds that are particularly strong .
thus droplets are expected to be non - convex with fractal surface of dimension @xmath75 , where @xmath76 is the space dimensionality .
the droplet model assumes that the energy of larger droplets is , on average , larger than that of smaller ones and hence that the spin glass phase is not destroyed by arbitrarily large , arbitrarily low energy , excitations .
in particular , it makes the scaling ansatz that the average free energies of these low lying excitations scales with their size , @xmath77 , @xmath78 . as a result of disorder , @xmath79 , and the distribution of @xmath80
is expected to be broad , with weight down to zero for all @xmath37 .
further the droplet model assumes that dynamics is activated and makes the second scaling assumption that activation barriers scale with droplet size , @xmath81 where the distribution of @xmath82 also has weight down to zero .
stiffness modulii @xmath83 and @xmath84 go to zero at the critical temperature @xmath85 as @xmath86 , where @xmath87 is the standard exponent .
feature ( i ) given above follows immediately from the assumptions of the scaling picture .
aging proceeds by equilibration of droplets in increasingly larger equilibrium domains of size @xmath88 . feature ( iii ) also follows immediately both from activated dynamics and from the temperature dependent stiffness coefficients .
thus the droplet picture predicts simple aging behavior as well as memory effects , at least under certain experimental conditions .
the dynamics provided by the droplet model of spin glasses differs importantly from the dynamics of a pure ferromagnet in the role of temperature change @xcite . a spin glass at @xmath89 is indeed described in terms of a renormalized hamiltonian with a renormalized ground state .
as mentioned above , this renormalized finite temperature hamiltonian is only statistically similar to the zero temperature hamiltonian .
hence the equilibrium states and droplet excitations thereof at various temperatures are not related to one another in any simple way .
in fact , a straightforward statistical argument @xcite provides an overlap length associated with a small temperature change from @xmath29 to @xmath90 , given by @xmath91 where @xmath92 .
the equilibrium state at @xmath29 is unchanged on scales @xmath93 but altered significantly on scales @xmath94 .
this ` temperature chaos ' realizes feature ( ii ) and predicts rejuvenation when @xmath95 is sufficiently small compared to @xmath96 @xcite .
remark that the droplet picture makes certain reasonable , but as yet unverified , assumptions .
it assumes that there are equilibrium states with reference to which low lying excitations can be defined , even as equilibrium domains are growing .
such assumptions are still a matter for debate @xcite . furthermore , the properties of these excitations are given by reasonable but assumed scaling laws . for work examining the validity these assumptions see refs .
@xcite .
even if the assumptions made in the droplet model are valid , it is not clear whether this model is sufficient to explain observed rejuvenation and memory effects . in order for it to do so
, the overlap length @xmath97 must be small enough to affect observations corresponding to experimental length scales .
we note that the characterization of @xmath97 in spin glasses is still a very active topic at the theoretical level @xcite .
a cautious conclusion is that , if @xmath98 exists , it certainly is too large to be observed in numerical simulations of systems like spin glasses .
however , because @xmath99 and @xmath82 are both broadly distributed , the droplet model anticipates that in rare regions , reorganization with change of temperature will take place even on unusually small length scales and experimentally viable time scales .
if such anomalous reorganization is common enough , rejuvenation and memory may still be accounted for by the droplet model even if the characteristic overlap length @xmath100 itself is not readily obtainable in experimental wait times @xcite .
the distribution of length scales about the overlap length @xmath100 certainly requires further theoretical investigation . on an experimental level
, rejuvenation effects do not constitute a proof of temperature chaos .
they are simply consistent with its existence .
in fact , in the simulational study of ref . @xcite it is argued that a rejuvenation effect is observed in a regime in which no chaotic effect of the kind described above can be detected @xcite .
hence , it is worth considering other mechanisms that might provide for rejuvenation and memory effects .
another mechanism has recently been invoked to account for rejuvenation effects , while retaining the simplicity of the domain growth approach @xcite .
we showed in section [ coarsening ] that the absence of rejuvenation in standard coarsening comes from the short - range character of thermal fluctuations . since the inequality @xmath41 prevents rejuvenation , the idea of ref .
@xcite is simply to consider opposite situations where @xmath101 .
away from critical points , @xmath102 is usually small .
the suggestion is hence to ` surf on a critical line ' @xcite where @xmath103 .
a simple aging experiments corresponds then to a quench to a critical point . in this case , the coherence length is equal to the dynamic correlation length @xmath104 , where @xmath105 is the dynamic critical exponent . aging is thus nothing but the successive equilibration of the critical fluctuations . similarities between this situation and aging in glassy materials were noted @xcite . when the temperature is shifted , all the critical fluctuations have to adapt to a new critical point with a new set of critical exponents .
this is the origin of the rejuvenation effect in this context .
it is now obvious that all the discussion of section [ back ] applies , with the dictionary `` objects '' = `` critical fluctuations '' and `` coherence length '' = `` dynamic correlation length '' .
the mechanism for rejuvenation is subtlely different from temperature chaos , even taken close to the critical point @xcite . here
, all length scales are always reshuffled without any ` overlap length ' @xcite .
the xy model is a microscopic realization of this picture .
it undergoes , in two dimensions , a kosterlitz - thouless transition from a standard paramagnetic state to a non - magnetized state with power - law correlations .
it presents then a ` line of critical points ' , with a continuous variation of the critical exponents with temperature . as a consequence ,
the 2d xy model exhibits aging , rejuvenation , memory and kovacs effects @xcite .
the example of a temperature cycle , where rejuvenation and memory are present , is given in fig .
[ surfing ] .
note that property ( ii ) is not naturally present in the model , since @xmath106 where @xmath107 is the ( renormalized ) spin wave stiffness . to build fig .
[ surfing ] , @xmath107 was arbitrarily , but not unreasonably , chosen to ensure the separation of length scales , @xmath108 .
this analogy raises the question : are glassy systems critical ?
a possible answer is yes they are , in which case they should have in common a line of critical points below the glass transition , in analogy with the 2d xy model .
a microscopic mechanism giving rise to such a behavior in , say , spin glasses , is lacking .
another is no ; but they do experimentally behave as if there is a critical line . the second , more pragmatic solution requires that @xmath36 is never decoupled from @xmath102 , even when the latter is finite .
one might argue this if the excitations on length scales exceeding @xmath102 are dominated by activation , and hence give exponentially increasing times for @xmath109 .
these conditions may make it virtually impossible to enter the regime @xmath110 .
slow evolution in the crossover region could result in effective critical behavior with continuous evolution of the exponents .
this is at least fully consistent with results in spin glasses @xcite .
we mention the anderson insulator , or ` electron glass ' , in this paper since its experimental properties were discussed at length in this school by ovadyahu @xcite .
hence , we do not present the system and directly discuss the results .
also , we only focus here on one of the many measurements performed on this system , and refer to refs .
@xcite for more details .
our aims are ( 1 ) to rephrase the `` aging experiment '' described by ovadyahu in the present context ; ( 2 ) to show that the phenomenology of section [ mini ] might be useful to interpret the data .
the electron glass is first prepared in its ` glassy phase ' .
this corresponds to a given low temperature and a gate voltage @xmath111 , used as the ` control parameter ' @xcite .
this corresponds then to a ` simple aging experiment ' .
aging manifests itself through a logarithmic time evolution of the conductance @xmath112 , used as a probe of the dynamics .
presumably , two - time functions would display in this regime interesting scaling behaviors , but no such measurements have been performed yet . after a very large time , typically days , @xmath112 is almost constant .
the gate voltage is then shifted to @xmath113 during a time @xmath114 .
experiments show that @xmath112 restarts to evolve in a logarithmic way .
this corresponds to a ` rejuvenation effect ' , since the days already spent in the glassy phase are apparently forgotten . in a third stage of the experiment ,
@xmath115 is shifted back to its initial value @xmath116 .
the experiments show that , after a time of order @xmath114 , @xmath112 reaches its initial value again .
this is analogous to what we called the ` transient ' , followed by the ` memory effect of the second kind ' .
furthermore , experiments have shown that during the third stage of the experiment ( ` transient ' and ` memory ' ) the conductance satisfies the scaling law @xmath117 where @xmath2 is the time counted from the beginning of the third stage @xcite .
this behavior was called ` aging ' in ref .
@xcite , because of the @xmath118 scaling , although the experiment is in fact a complete cycle in terms of @xmath115 .
can one go beyond words and analogies ? if the similarity between these experiments and standard temperature cycling experiments is assumed , then the cartoon of fig .
[ cartoon ] can be used .
the @xmath118 scaling of the conductance tells that the duration of the ` transient ' has the same magnitude as the duration of the second stage .
this means that there is almost no separation of time scales in the electron glass for this range of gate voltages , as noted in ref .
building further on analogies , one can reproduce this experiment using the 2d xy model .
stay first a very long ( infinite ) time at temperature @xmath11 , then do a shift of duration @xmath114 at @xmath14 and come back at @xmath11 at time @xmath10 . in this case
, one can compute the time dependence of the spin autocorrelation function between times @xmath2 and @xmath119 , noted @xmath120 , as a function of the duration of the shift @xmath114 .
one finds , dropping irrelevant factors and using the notations of section [ xy ] : @xmath121 where @xmath122 is the equilibrium correlation function obtained when @xmath123 ( no shift at all ) and @xmath124 and @xmath125 are the usual critical exponents at temperatures @xmath11 and @xmath14 .
the second and third term of the right hand side respectively account for the ` transient ' and the ` memory ' . for @xmath126 ,
the transient term is dominant and aging is observed through a power law decay . when @xmath127 , both terms combine to restore equilibrium , accounting for the memory .
interestingly , the crossover time scale @xmath128 is given by @xmath129 no separation of length scales amounts thus to @xmath130 as is observed in the electron glass .
deviations from @xmath118 , as observed at higher gate voltages @xmath115 @xcite , could be tentatively related via eq .
( [ equation ] ) to a decrease of an activation energy at high @xmath115 .
this effect could be more systematically investigated using eqs .
( [ equation ] ) and ( [ cross ] ) .
overaging was introduced in section [ phenomeno ] .
it is reproduced in fig .
[ overaging ] showing a multiple light scattering experiment on a colloidal suspension @xcite .
two interesting points in this experiment are ( 1 ) the control parameter is an external oscillatory shear strain , as opposed to temperature in ` standard ' shift experiments ; ( 2 ) no obvious coherence length is known to grow during the aging of colloids . therefore , such experiments could possibly be used to characterize length scales in colloidal glasses . in this context ,
` simple aging experiments ' consist in a ` quench ' from a shear strain of very large amplitude ( typically more than 20 % ) mimicking a ` high temperature ' .
standard @xmath33 scalings have been observed in many similar soft glassy materials @xcite .
it is thus natural to start and investigate more complex protocols , in the spirit of shifts and cycles @xcite .
a first possibility is to quench first towards a small , but finite , shear strain during a time @xmath114 , after which the shear is completely stopped .
one can then compare the results with a temperature shift experiment , @xmath131 , which are more commonly performed . in spin glasses , in agreement with the phenomenology developed above , the following results are known .
we already discussed the case of large @xmath132 : rejuvenation is observed @xcite . for very small
@xmath133 , the relaxation of the system after the shift has the same shape as in a direct quench to @xmath14 , but with an ` effective age ' @xmath134 @xcite .
indeed , a small @xmath133 means that the objects growing at @xmath11 will be almost unchanged at @xmath14 ( no rejuvenation ) .
however , due the property ( iii ) , the time spent at higher temperature is more efficient in growing the coherence length . hence , the relaxation is slower after the shift , the system looking ` older ' or ` overaged '
this is observed in the colloidal suspension in a ` strain - shift ' experiment , see fig .
[ overaging](a ) .
more quantitatively , the effective age @xmath135 can be estimated in terms of the coherence length as @xcite @xmath136 repeating this protocol for various ( @xmath137 gives thus quantitative informations on the growth law of the coherence length @xcite .
furthermore , eq .
( [ equation ] ) shows that this protocol allows a quantitative study of the energy barriers encountered by the system during its nonequilibrium dynamics @xcite . for intermediate @xmath133 , of course
, the result will be a combination of rejuvenation and overaging .
this has been observed in spin glasses @xcite and in polymers @xcite . that this is the case in the colloidal suspension with oscillatory strain as a control parameter
is demonstrated in fig .
[ overaging](b ) . again
, systematic studies using eqs . ( [ equation ] ) and ( [ tweff ] ) could possibly lead to a quantitative characterization of a coherence length in this and other systems .
this paper is the result of a discussion group organized during this summer school . it has become a kind of a review of what glassy dynamics looks like , as encountered in experiments performed on different materials using diverse control parameters .
although first observed in spin and structural glasses , it now is clear that these spectacular effects are quite generic .
this points towards the necessity of having a simple interpretation of these phenomena and we have used the concept of a coherence length to develop one .
note that all the theoretical interpretations discussed in this paper rely on a some sort of hierarchical picture : hierarchy of traps levels , of time scales in infinite - range glass models , of length scales in spatial approaches .
we have focused on the latter because we think it gives greater physical insight .
we note , however , that definition , characterization and measurement of a coherence length is still largely an open problem for most of the systems discussed here ! in this respect , we mentioned several times that the experimental protocols discussed in this paper and their interpretation , for instance eq .
( [ cross ] ) and eq .
( [ tweff ] ) , constitute a promising starting point for progress on the crucial problem of length scales in glassy materials .
we would like to express our gratitioude to the organizers for giving us the opportunity to appear in these proceedings .
we also acknowledge discussions and collaborations with g. biroli , j .-
bouchaud , d. fisher , p. holdsworth , r. da silveira , h. yoshino , and p. young .
m. kroon , g. wegdam and r. sprik , phys .
e * 54 * , 1 ( 1996 ) ; d. bonn , h. tanaka , g. wegdam , h. kellay and j. meunier , europhys .
* 45 * , 52 ( 1999 ) ; a. knaebel , m. bellour , j .-
munch , v. viasnoff , f. lequeux and j. l. harden , europhys . lett . * 52 * , 73 ( 2000 ) .
j. b. knight , c. g. fandrich , c. n. lau , h. m. jaeger and s. r. nagel , phys .
e * 51 * , 3957 ( 1995 ) ; c. josserand , a. tkachenko , d. m. mueth and h. m. jaeger , phys .
85 * , 3632 ( 2000 ) ; m. nicolas , p. duru and o. pouliquen , eur .
j. e * 3 * , 309 ( 2000 ) ; p.
philippe and d. bideau , europhys . lett .
( in press ) , preprint cond - mat/0210297 .
e. vincent , j. hamman , m. ocio , j .-
bouchaud and l. f. cugliandolo , in _ complex behavior of glassy systems _ , ed . : m. rubi ( springer verlag , berlin , 1997 ) ; p. nordblad and p. svendlidh , in _ spin glasses and random fields _ , ed .
young ( world scientific , singapore , 1998 ) .
g. grinstein and s. k. ma , phys .
b * 28 * , 2588 ( 1983 ) ; g. grinstein and j. f. fernandez , phys . rev .
b * 29 * , 6389 ( 1984 ) ; d. a. huse and c. henley , phys .
lett . * 54 * , 2708 ( 1985 ) ; d. s. fisher and a. a. middleton , preprint cond - mat/0107489 .
a. j. bray and m. a. moore , j. phys .
c * 17 * , l463 ( 1984 ) ; and in _ heidelberg colloquim on glassy dynamics _ , lectures notes in physics * 275 * , eds . :
j. l. van hemmen and i. morgenstern ( springer , berlin , 1987 ) . d. a. huse and d. s. fisher , j.phys .
a * 20 * , l997 ( 1987 ) ; d. s. fisher and d. a. huse , j.phys a * 20 * , l1005 ( 1987 ) ; c. m. newman and d. l. stein , phys .
b * 46 * , 973 ( 1992 ) ; phys . rev
. lett . * 72 * , 2286 ( 1994 ) ; _ ibid . _ * 76 * , 515 ( 1996 ) ; _ ibid . _ * 76 * , 4821 ( 1996 ) ; _ ibid . _ * 84 * , 3966 ( 2000 ) ; phys .
e * 55 * , 5194 ( 1997 ) ; _ ibid . _ * 57 * , 1356 ( 1998 ) ; _ ibid .
_ * 63 * , 160101 ( 2001 ) ; a. a. middleton , phys . rev
. lett . * 83 * , 1672 ( 1999 ) ; e. marinari , g. parisi , f. ricci - tersenghi , j. ruiz - lorenzo , and f. zuliani , j. stat . phys . * 98 * , 973 ( 2000 ) ; d. s. fisher , this volume ; f. krzakala and o. c. martin , phys . rev
. lett . * 85 * , 3017 ( 2000 ) ; m. palassini and a. p.
young , phys .
lett . * 85 * , 3017 ( 2000 ) .
j. lamarcq , j .-
bouchaud , o. c. martin and m. mzard , europhys . lett .
* 58 * , 321 ( 2002 ) ; a. a. middleton , phys .
b * 63 * , 060202 ( 2001 ) ; j. houdayer , f. krzakala , and o. c. martin , eur .
j. b * 18 * , 467 ( 2000 ) .
t. aspelmeier , a. j. bray , and m. a. moore , phys .
89 * , 197202 ( 2002 ) ; t. rizzo and a. crisanti , preprint cond - mat/0209333 ; m. sales and h. yoshino , phys .
e * 65 * , 066131 ( 2002 ) ; d. a. huse and l. f. ko , phys .
b * 56 * , 14597 ( 1997 ) ; f. krzakala and o. c. martin eur .
j. b * 28 * , 199 ( 2002 ) .
although ` surfing on glassy phases ' would have been a possible title for this paper , we believe that ` hiking ' is better suited in the context of a summer school taking place so close to the alps and the ` mer de glace ' . | experiments performed on a wide range of glassy materials display many interesting phenomena , such as aging behavior . in recent years , a large body of experiments probed this nonequilibrium glassy dynamics through elaborate protocols , in which external parameters are shifted , or cycled in the course of the experiment .
we review here these protocols , as well as experimental and numerical results .
then , we critically discuss various theoretical approaches put forward in this context .
emphasis is put more on the generality of the phenomena than on a specific system .
experiments are also suggested . |
we address the problem of clustering words , or that of constructing a thesaurus , based on co - occurrence data .
we view this problem as that of estimating a joint probability distribution over word pairs , specifying the joint probabilities of word pairs , such as noun verb pairs . in this paper
, we assume that the joint distribution can be expressed in the following manner , which is stated for noun verb pairs for the sake of readability : the joint probability of a noun and a verb is expressed as the product of the joint probability of the noun class and the verb class which the noun and the verb respectively belong to , and the conditional probabilities of the noun and the verb given their respective classes . as a method for estimating such a probability distribution
, we propose an algorithm based on the minimum description length ( mdl ) principle .
our clustering algorithm iteratively merges noun classes and verb classes in turn , in a bottom up fashion . for each merge it performs
, it calculates the increase in data description length resulting from merging any noun ( or verb ) class pair , and performs the merge having the least increase in data description length , provided that the increase in data description length is less than the reduction in model description length .
there have been a number of methods proposed in the literature to address the word clustering problem ( e.g. , @xcite ) .
the method proposed in this paper is a natural extension of both li & abe s and brown et al s methods , and is an attempt to overcome their drawbacks while retaining their advantages .
the method of brown et al , which is based on the maximum likelihood estimation ( mle ) , performs a merge which would result in the least reduction in ( average ) mutual information .
our method turns out to be equivalent to performing the merge with the least reduction in mutual information , provided that the reduction is below a certain threshold which depends on the size of the co - occurrence data and the number of classes in the current situation .
this method , based on the mdl principle , takes into account both the fit to data and the simplicity of a model , and thus can help cope with the over - fitting problem that the mle - based method of brown et al faces . the model employed in @xcite
is based on the assumption that the word distribution within a class is a uniform distribution , i.e. every word in a same class is generated with an equal probability .
employing such a model has the undesirable tendency of classifying into different classes those words that have similar co - occurrence patterns but have different absolute frequencies .
the proposed method , in contrast , employs a model in which different words within a same class can have different conditional generation probabilities , and thus can classify words in a way that is not affected by words absolute frequencies and resolve the problem faced by the method of @xcite .
we evaluate our clustering method by using the word classes and the joint probabilities obtained by it in syntactic disambiguation experiments .
our experimental results indicate that using the word classes constructed by our method gives better disambiguation results than when using li & abe or brown et al s methods . by combining thesauruses automatically constructed by our method and an existing hand - made thesaurus ( wordnet )
, we were able to achieve the overall accuracy of @xmath0 for pp - attachment disambiguation , which compares favorably against the accuracy ( @xmath1 ) obtained using the state - of - the - art method of @xcite .
suppose available to us are co - occurrence data over two sets of words , such as the sample of verbs and the head words of their direct objects given in fig .
[ fig : co - occurrence ] .
our goal is to ( hierarchically ) cluster the two sets of words so that words having similar co - occurrence patterns are classified in the same class , and output a thesaurus for each set of words . we can view this problem as that of estimating the best probability model from among a class of models of ( probability distributions ) which can give rise to the co - occurrence data . in this paper , we consider the following type of probability models .
assume without loss of generality that the two sets of words are a set of nouns @xmath2 and a set of verbs @xmath3 .
a partition @xmath4 of @xmath2 is a set of noun - classes satisfying @xmath5 and @xmath6 .
a partition @xmath7 of @xmath3 can be defined analogously .
we then define a probability model of noun - verb co - occurrence by defining the joint probability of a noun @xmath8 and a verb @xmath9 as the product of the joint probability of the noun and verb classes that @xmath8 and @xmath9 belong to , and the conditional probabilities of @xmath8 and @xmath9 _ given _ their classes , that is , @xmath10 where @xmath11 and @xmath12 denote the ( unique ) classes to which @xmath8 and @xmath9 belong . in this paper
, we refer to this model as the ` hard clustering model , ' since it is based on a type of clustering in which each word can belong to only one class .
[ fig : hardmodel ] shows an example of the hard clustering model that can give rise to the co - occurrence data in fig .
[ fig : co - occurrence ] .
a particular choice of partitions for a hard clustering model is referred to as a ` discrete ' hard - clustering model , with the probability parameters left to be estimated .
the values of these parameters can be estimated based on the co - occurrence data by the maximum likelihood estimation . for a given set of co - occurrence data
@xmath13 the maximum likelihood estimates of the parameters are defined as the values that maximize the following likelihood function with respect to the data : @xmath14 it is easy to see that this is possible by setting the parameters as @xmath15 @xmath16 here , @xmath17 denotes the entire data size , @xmath18 the frequency of word pairs in class pair @xmath19 , @xmath20 the frequency of word @xmath21 , and @xmath22 the frequency of words in class @xmath23 .
the question now is what criterion should we employ to select the best model from among the possible models . here
we adopt the minimum description length ( mdl ) principle .
mdl @xcite is a criterion for data compression and statistical estimation proposed in information theory . in applying mdl
, we calculate the code length for encoding each model , referred to as the ` model description length ' @xmath24 , the code length for encoding the given data through the model , referred to as the ` data description length ' @xmath25 and their sum : @xmath26 the mdl principle stipulates that , for both data compression and statistical estimation , the best probability model with respect to _ given data _ is that which requires the least total description length . the data description length is calculated as @xmath27 where @xmath28 stands for the maximum likelihood estimate of @xmath29 ( as defined in section [ sec : estimate ] ) .
we then calculate the model description length as @xmath30 where @xmath31 denotes the number of _ free _ parameters in the model , and @xmath17 the entire data size .
, and the conditional probabilities @xmath32 and @xmath33 .
since these alternatives are approximations of one another asymptotically , here we use only the simplest formulation . in the full paper
, we plan to compare the empirical behavior of the alternatives . ] in this paper , we ignore the code length for encoding a ` discrete model , ' assuming implicitly that they are equal for all models and consider only the description length for encoding the parameters of a model as the model description length .
if computation time were of no concern , we could in principle calculate the total description length for each model and select the optimal model in terms of mdl . since the number of hard clustering models is of order @xmath34 , where @xmath35 and @xmath36 denote the size of the noun set and the verb set , respectively , it would be infeasible to do so . we therefore need to devise an efficient algorithm that heuristically performs this task .
the proposed algorithm , which we call ` 2d - clustering , ' iteratively selects a suboptimal mdl - model from among those hard clustering models which can be obtained from the current model by merging a noun ( or verb ) class pair .
as it turns out , the minimum description length criterion can be reformalized in terms of ( average ) mutual information , and a greedy heuristic algorithm can be formulated to calculate , in each iteration , the reduction of mutual information which would result from merging any noun ( or verb ) class pair , and perform the merge having the least mutual information reduction , _ provided that the reduction is below a variable threshold_. 2d - clustering@xmath37 + ( @xmath38 is the input co - occurrence data , and @xmath39 and @xmath40 are positive integers . ) 1 .
initialize the set of noun classes @xmath4 and the set of verb classes @xmath7 as : @xmath41 where @xmath2 and @xmath3 denote the noun set and the verb set , respectively .
2 . repeat the following three steps : 1 . execute merge@xmath42 to update @xmath4 , 2 .
execute merge@xmath43 to update @xmath7 , 3 .
if @xmath4 and @xmath7 are unchanged , go to step 3 .
3 . construct and output a thesaurus for nouns based on the history of @xmath4 , and one for verbs based on the history of @xmath7 .
next , we describe the procedure of ` merge , ' as it is being applied to the set of noun classes with the set of verb classes fixed .
merge@xmath42 1 . for each class pair in @xmath4 ,
calculate the reduction of mutual information which would result from merging them .
( the details will follow . )
discard those class pairs whose mutual information reduction ( [ eq : implement ] ) is not less than the threshold of @xmath44 where @xmath17 denotes the total data size , @xmath45 the number of free parameters in the model before the merge , and @xmath46 the number of free parameters in the model after the merge .
sort the remaining class pairs in ascending order with respect to mutual information reduction .
2 . merge the first @xmath39 class pairs in the sorted list .
3 . output current @xmath4 .
we perform ( maximum of ) @xmath39 merges at step 2 for improving efficiency , which will result in outputting an at - most @xmath39-ary tree .
note that , strictly speaking , once we perform one merge , the model will change and there will no longer be a guarantee that the remaining merges still remain justifiable from the viewpoint of mdl .
next , we explain why the criterion in terms of description length can be reformalized in terms of mutual information .
we denote the model before a merge as @xmath47 and the model after the merge as @xmath48 . according to mdl
, @xmath48 should have the least increase in data description length @xmath49 and at the same time satisfies @xmath50 this is due to the fact that the decrease in model description length equals @xmath51 and is identical for each merge .
in addition , suppose that @xmath48 is obtained by merging two noun classes @xmath52 and @xmath53 in @xmath47 to a single noun class @xmath54 .
we in fact need only calculate the difference between description lengths with respect to these classes , i.e. , @xmath55 now using the identity @xmath56 we can rewrite the above as @xmath57 thus , the quantity @xmath58 is equivalent to the mutual information reduction times the data size . and
@xmath7 is defined as @xmath59 ] we conclude therefore that in our present context , a clustering with the least data description length increase is equivalent to that with the least mutual information decrease .
canceling out @xmath60 and replacing the probabilities with their maximum likelihood estimates , we obtain @xmath61 therefore , we need calculate only this quantity for each possible merge at step 1 of merge . in our implementation of the algorithm
, we first load the co - occurrence data into a matrix , with nouns corresponding to rows , verbs to columns . when merging a noun class in row @xmath62 and that in row @xmath63 ( @xmath64 ) , for each @xmath12 we add @xmath65 and @xmath66 obtaining @xmath67 , write @xmath67 on row @xmath62 , move @xmath68 to row @xmath63 , and reduce the matrix by one row . by the above implementation
, the worst case time complexity of the algorithm is @xmath69 where @xmath35 denotes the size of the noun set , @xmath36 that of the verb set .
if we can merge @xmath39 and @xmath40 classes at each step , the algorithm will become slightly more efficient with the time complexity of @xmath70 .
we can restrict the hard clustering model ( [ eq : hardmodel ] ) by assuming that words within a same class are generated with an equal probability , obtaining @xmath71 which is equivalent to the model proposed by @xcite . employing this restricted model
has the undesirable tendency to classify into different classes those words that have similar co - occurrence patterns but have different absolute frequencies .
the hard clustering model defined in ( [ eq : hardmodel ] ) can also be considered to be an extension of the model proposed by brown et al .
first , dividing ( [ eq : hardmodel ] ) by @xmath72 , we obtain @xmath73 since hard clustering implies @xmath74 holds , we have @xmath75 in this way , the hard clustering model turns out to be a class - based bigram model and is similar to brown et al s model .
the difference is that the model of ( [ eq : hardmodel2 ] ) assumes that the clustering for @xmath11 and the clustering for @xmath12 can be different , while the model of brown et al assumes that they are the same .
a very general model of noun verb joint probabilities is a model of the following form : @xmath76 here @xmath77 denotes a set of noun classes satisfying @xmath78 , but not necessarily disjoint . similarly @xmath79 is a set of not necessarily disjoint verb classes .
we can view the problem of clustering words in general as estimation of such a model .
this type of clustering in which a word can belong to several different classes is generally referred to as ` soft clustering . '
if we assume in the above model that each verb forms a verb class by itself , then ( [ eq : softmodel ] ) becomes @xmath80 which is equivalent to the model of pereira et al . on the other hand ,
if we restrict the general model of ( [ eq : softmodel ] ) so that both noun classes and verb classes are disjoint , then we obtain the hard clustering model we propose here ( [ eq : hardmodel ] ) .
all of these models , therefore , are some special cases of ( [ eq : softmodel ] ) .
each specialization comes with its merit and demerit .
for example , employing a model of soft clustering will make the clustering process more flexible but also make the learning process more computationally demanding .
our choice of hard clustering obviously has the merits and demerits of the soft clustering model reversed .
our method is also an extension of that proposed by brown et al from the viewpoint of estimation criterion .
their method merges word classes so that the reduction in mutual information , or equivalently the increase in data description length , is minimized .
their method has the tendency to overfit the training data , since it is based on mle .
employing mdl can help solve this problem .
we apply the acquired word classes , or more specifically the probability model of co - occurrence , to the problem of structural disambiguation . in particular , we consider the problem of resolving pp - attachment ambiguities in quadruples , like ( see , girl , with , telescope ) and that of resolving ambiguities in compound noun triples , like ( data , base , system ) . in the former ,
we determine to which of ` see ' or ` girl ' the phrase ` with telescope ' should be attached . in the latter
, we judge to which of ` base ' or ` system ' the word ` data ' should be attached .
we can perform pp - attachment disambiguation by comparing the probabilities @xmath81 if the former is larger , we attach ` with telescope ' to ` see ; ' if the latter is larger we attach it to ` girl ; ' otherwise we make no decision .
( disambiguation on compound noun triples can be performed similarly . ) since the number of probabilities to be estimated is extremely large , estimating all of these probabilities accurately is generally infeasible ( i.e. , the data sparseness problem ) . using our clustering model to calculate these conditional probabilities ( by normalizing the joint probabilities with marginal probabilities ) can solve this problem .
we further enhance our disambiguation method by the following back - off procedure : we first estimate the two probabilities in question using hard clustering models constructed by our method .
we also estimate the probabilities using an existing ( hand - made ) thesaurus with the ` tree cut ' estimation method of @xcite , and use these probability values when the probabilities estimated based on hard clustering models are both zero .
finally , if both of them are still zero , we make a default decision .
in this experiment , we used heuristic rules to extract verbs and the head words of their direct objects from the _ tagged _ texts of the wsj corpus ( acl / dci cd - rom1 ) consisting of 126,084 sentences .
we then constructed a number of thesauruses based on these data , using our method .
[ fig : exthesau ] shows a part of a thesaurus for 100 randomly selected nouns , based on their appearances as direct objects of 20 randomly selected verbs .
the thesaurus seems to agree with human intuition to some degree , although it is constructed based on a relatively small amount of co - occurrence data .
for example , ` stock , ' ` security , ' and ` bond ' are classified together , despite the fact that their absolute frequencies in the data vary a great deal ( 272 , 59 , and 79 , respectively . )
the results demonstrate a desirable feature of our method , namely , it classifies words based solely on the similarities in co - occurrence data , and is not affected by the absolute frequencies of the words .
we extracted compound noun doubles ( e.g. , ` data base ' ) from the tagged texts of the wsj corpus and used them as training data , and then conducted structural disambiguation on compound noun triples ( e.g. , ` data base system ' ) .
we first randomly selected 1,000 nouns from the corpus , and extracted compound noun doubles containing those nouns as training data and compound noun triples containing those nouns as test data . there were 8,604 training data and 299 test data .
we hand - labeled the test data with the correct disambiguation ` answers . '
we performed clustering on the nouns on the left position and the nouns on the right position in the training data by using both our method ( ` 2d - clustering ' ) and brown et al s method ( ` brown ' ) .
we actually implemented an extended version of their method , which separately conducts clustering for nouns on the left and those on the right ( which should only improve the performance ) .
4.5 cm we next conducted structural disambiguation on the test data , using the probabilities estimated based on 2d - clustering and brown .
we also tested the method of using the probabilities estimated based on word co - occurrences , denoted as ` word - based . ' fig .
[ fig : compound ] shows the results in terms of accuracy and coverage , where coverage refers to the percentage of test data for which the disambiguation method was able to make a decision . since for brown the number of classes finally created has to be designed in advance , we tried a number of alternatives and obtained results for each of them .
( note that , for 2d - clustering , the optimal number of classes is automatically selected . )
[ tab : compound ] .compound noun disambiguation results [ cols="<,^",options="header " , ] we then enhanced each of these methods by using a default rule when a decision can not be made , which is indicated as ` + default . '
[ tab : pp - attach - final ] shows the results of these experiments .
we can make a number of observations from these results .
( 1 ) 2d - clustering achieves a broader coverage than nounclass-2dc . this is because in order to estimate the probabilities for disambiguation , the former exploits more information than the latter .
( 2 ) for brown , we show here only its best result , which happens to be the same as the result for 2d - clustering , but in order to obtain this result we had to take the trouble of conducting a number of tests to find the best level of clustering . for 2d - clustering , this was done once and automatically . compared with li - abe96
, 2d - clustering clearly performs better .
therefore we conclude that our method improves these previous clustering methods in one way or another .
( 3 ) 2d - clustering outperforms wordnet in term of accuracy , but not in terms of coverage .
this seems reasonable , since an automatically constructed thesaurus is more domain dependent and therefore captures the domain dependent features better , and thus can help achieve higher accuracy . on the other hand , with the relatively small size of training data we had available , its coverage is smaller than that of a general purpose hand made thesaurus .
the result indicates that it makes sense to combine automatically constructed thesauruses and a hand - made thesaurus , as we have proposed in section [ sec : disam ] .
this method of combining both types of thesauruses ` 2d - clustering+wordnet+default ' was then tested .
we see that this method performs the best .
( see tab .
[ tab : pp - attach - final ] . ) finally , for comparison , we tested the ` transformation - based error - driven learning ' proposed in @xcite , which is a state - of - the - art method for pp - attachment disambiguation . tab .
[ tab : pp - attach - final ] shows the result for this method as ` brill - resnik . '
we see that our disambiguation method also performs better than brill - resnik .
( note further that for brill & resnik s method , we need to use quadruples as training data , whereas ours only requires triples . )
we have proposed a new method of clustering words based on co - occurrence data .
our method employs a probability model which naturally represents co - occurrence patterns over word pairs , and makes use of an efficient estimation algorithm based on the mdl principle .
our clustering method improves upon the previous methods proposed by brown et al and @xcite , and furthermore it can be used to derive a disambiguation method with overall disambiguation accuracy of @xmath0 , which improves the performance of a state - of - the - art disambiguation method .
the proposed algorithm , 2d - clustering , can be used in practice , as long as the data size is at the level of the current penn tree bank . yet
it is still relatively computationally demanding , and thus an important future task is to further improve on its computational efficiency .
we are grateful to dr .
s. doi of nec c&c media res .
labs . for his encouragement .
we thank ms .
y. yamaguchi of nis for her programming efforts . | we address the problem of clustering words ( or constructing a thesaurus ) based on co - occurrence data , and using the acquired word classes to improve the accuracy of syntactic disambiguation .
we view this problem as that of estimating a joint probability distribution specifying the joint probabilities of word pairs , such as noun verb pairs .
we propose an efficient algorithm based on the minimum description length ( mdl ) principle for estimating such a probability distribution .
our method is a natural extension of those proposed in @xcite and @xcite , and overcomes their drawbacks while retaining their advantages .
we then combined this clustering method with the disambiguation method of @xcite to derive a disambiguation method that makes use of both automatically constructed thesauruses and a hand - made thesaurus .
the overall disambiguation accuracy achieved by our method is @xmath0 , which compares favorably against the accuracy ( @xmath1 ) obtained by the state - of - the - art disambiguation method of @xcite . |
in recent years there have been several studies of the cosmological history of star formation rate ( sfr ) in galaxies .
usually the sfr is derived from the luminosity density in a selected wavelength band which is an indicator of star formation .
some of the sf indicators that have been used are : h@xmath0 line , far - infrared ( fir ) , submm , ultraviolet ( uv ) and nonthermal radio continuum emissions .
each indicator is sensitive to a certain range of stellar masses and the total sfr is obtained by assuming an initial mass function ( imf ) of stars at birth , usually a power - law form .
further , the observed luminosity depends on the star formation history and the epoch of observation . even for a constant rate of star formation
, there is a time evolution of luminosity which increases , reaches an equilibrium value and declines after the termination of sf .
this evolution and decline also depend on the exponent of the imf .
the changes in the luminosity to sfr ratio as a function of these parameters are not the same for all indicators .
thus , apart from the observational errors , uncertainties arise in the sfr derived from luminosity density using different indicators . in this paper
, we study these uncertainties by computing the evolution of the luminosity to sfr ratio for different imfs and scenarios of star formation .
it should be noted that there are other parameters like metallicity that can affect the derivation of sfr from luminosity .
for example , hirashita et al ( 2001 ) find that the fir - sfr conversion factor can vary by a factor of up to 4 on this account .
when we have better knowledge of factors like metallicity and cirrus contribution in starburst galaxies , it may be possible to quantify these uncertainties .
in this study we compute the time evolution of l / sfr ratio for the following luminosities : l@xmath1 , the bolometric luminosity , l@xmath2 , the far - infrared luminosity which is a fraction of the bolometric luminosity , l@xmath3 , the intrinsic luminosity in an ultraviolet band , l@xmath4 , the luminosity of uv photons escaping the star forming region , l@xmath5 , the h@xmath0 line luminosity and l@xmath6 , the nonthermal radio continuum luminosity . for uv
we use the 0.16 @xmath7 m band which has been extensively used in hst observations .
the submm luminosity is usually converted into a fir luminosity to derive the sfr and hence we do not consider it separately . the radio continuum is assumed to be from synchrotron emission from cosmic ray electrons accelerated by supernovae arising from stars of mass @xmath8 8 m@xmath9 . in starburst(sb ) galaxies , which are the subjects of our study , the contribution of the earlier disc population is small and is neglected .
their contribution , if present , will add additional uncertainties . the fir luminosity is dust - reradiated emission and in many studies , it has been taken to be the same as the bolometric luminosity .
though this is likely to be the case at the very early embedded stage of a star , the efficiency of conversion will be less than unity and decrease as time passes because of the disruption and dispersion of the parent clouds by the massive stars formed .
the fact that visible ob stars are observed in star forming complexes in the milky way supports this hypothesis .
silva et al . (
1998 ) have introduced the concept of residence time and have shown that their models give good fits to observational spectral energy distributions and spectral line data . in this study , we assume that the fir conversion efficiency decreases exponentially with a time constant @xmath10 .
it then follows that the flux of uv photons escaping the star forming region depends inversely on this conversion efficiency and can be characterized by an extinction which decreases exponentially with the same time constant .
the h@xmath0 luminosity will also have a similar dependence with a proper scaling of the extinction .
the total luminosity in a given band is the integral over the history of stellar formation and evolution and over the stellar masses .
the imf is assumed to have a power law form @xmath11(m ) @xmath12 m@xmath13 within mass limits of 120 to 1 m@xmath9 . since the contribution of stars of mass @xmath14 1 m@xmath9 to the luminosity , in the bands considered , is negligible , the l / sfr ratio can easily be scaled for an imf extending to lower masses .
the stellar evolutionary tracks are taken from schaller et al .
( 1992 ) and the computational procedure follows that of mayya ( 1995 ) . the integrated luminosity at time t , the epoch of observation is given by @xmath15 @xmath16 @xmath17 @xmath18 @xmath19 @xmath20 here t is the look back time when a star is born , t is the observational epoch from a reference time when the starburst started , p(t ) is the production rate of stars at time t , t@xmath21(m ) is the lifetime of a star of mass m , a(uv ) and a(h ) are the extinctions in magnitudes at 0.16 @xmath7 m and h@xmath0 wavelength at the time of birth , lyc refers to lyman continuum photons and k@xmath5 is a constant that relates h@xmath0 luminosity to the lyman continuum luminosity and k@xmath6 is the nonthermal radio continuum luminosity per supernova and @xmath22 is the exponential residence time of synchrotron emitting electrons .
the production term p(t ) represents the time evolution of the starburst .
it is a @xmath23 function for an instantaneous burst , constant for a continuous and constant star formation and could have other forms like an exponential decay or a gaussian .
the imf index is neither known accurately nor is it known to be universal .
observations of stellar clusters in the milky way and the magellanic clouds give a range of 2 to 3 ( scalo 1998 ) .
we , therefore , perform computations for three values of @xmath24 = 2 , 2.5 and 3 . as for the time dependence of star formation
, we will consider , in this paper , only the case of p(t ) = constant for t = 0 to t@xmath25 and zero elsewhere and present the results for @xmath26 = 95 myr . as for the residence time of stars in the parent clouds ,
silva et al . ( 1998 ) find that for the starburst galaxies they model , a value of 1020 myr fits the observations .
here , we take @xmath10 = 15 myr . the extinction at 0.16 @xmath7 m and h@xmath0 are taken as 10 and 3 magnitudes respectively .
the uv extinction adopted leads to a luminosity weighted extinction factor of 5 - 6 at times @xmath8 50 myr similar to the value estimated by steidel et al .
it may be noted that for moderately obscured regions , a more elaborate attenuation model is needed .
however , for the luminous starburst galaxies , especially at moderate and large redshifts , our treatment may be adequate . as remarked earlier the presence of older star - forming regions will add further uncertainty to the sfr .
for the residence time of high energy electrons , we take @xmath22 = 10 myr .
in figure 1 , we show for different sf indicators , the time evolution of the ratio of luminosity at time t to that at 95 myr when the sb terminates . we have not shown the curve for h@xmath0 since it is very similar to that of the fir . for fir and uve
, we show the ratio with respect to the bolometric and intrinsic uv luminosity respectively .
the general trend of the time evolution is an increase of the luminosity to a plateau value and then a decrease after the sb terminates.the plateau values of the ratio of luminosity to sfr for the bolometric , fir and uv ( 0.16 @xmath7m)emissions are 1.6@xmath2710@xmath28 l@xmath9m@xmath9@xmath29 yr , 8.8@xmath2710@xmath30 l@xmath9m@xmath9@xmath29 yr and 5.5@xmath2710@xmath31 w hz@xmath29 m@xmath9@xmath29 yr respectively . these values , after adjustment for the salpeter imf and the 0.1100 m@xmath9 range adopted by kennicutt ( 1998),are about the same as those given in his review .
further , the ratio of fir to radio continuum fluxes is similar to that of bressan et al ( 2002 ) if we adopt their value of nonthermal emission per supernova .
= 2.5 . for fir and uve , escaping uv luminosity , ratio with respect to bolometric and intrinsic uv luminosity respectively are shown with the same numerical scale.,width=377 ] = 2.5 plotted against @xmath24 for three observational epochs .
the points are joined together to facilitate viewing.,width=377 ] the rise- and fall - times vary from band to band .
the time profile for the fir luminosity is the sharpest while it is the broadest for the uve .
this is due to the fact that luminous massive stars have lifetimes @xmath32 the residence time and almost the total bolometric luminosity is re - radiated in the fir while for stars with lifetime @xmath8 15 myr , the conversion efficiency decreases .
the opposite is the case for the escaping uv photons .
for @xmath10 = 15 myr , the maximum conversion efficiency is 0.55
. it will be lower if @xmath10 is less .
the fraction of escaping uv photons reaches a maximum of 0.17 .
the uv extinction is more sensitive to the values of @xmath10 and extinction at zero age than the fir conversion efficiency .
the slight increase in the uve luminosity after the termination of the sb and the subsequent oscillation is due to a competition between the death of massive stars and increased escape of uv photons from lower mass stars as the stars age .
it may be noted that the fir profile is sharper than that of the bolometric luminosity while the uve approaches that of uvi in the post starburst period .
the radio continuum luminosity has a rise- and fall - time of about 30 myr , the lifetime of stars of mass 8 m@xmath9 . the plateau fir and uve luminosities depend on @xmath10 , the residence time .
hence , there will be additional uncertainties in the sfr resulting from the variation of @xmath10 from galaxy to galaxy and its possible dependence on the intensity of starburst .
it is generally believed that the fir emission is least subject to extinction correction applicable to other wavelengths . however , the concept of residence time introduces corrections up to a factor of 23 . from this point of view , the radio continuum emission is a better probe of sfr since it needs no such correction .
it is clear that the sfr derived for an individual galaxy depends on the duration of the sb and the time of observation and the luminosity does not give the instantaneous current sfr unless one observes in the plateau region of continuous and constant star formation . even if we observe a large enough sample of galaxies , say in a given redshift bin , and all galaxies have the same duration of starburst and other parameters remain the same , the uncertainty is not fully eliminated .
if we assume that the sb duration is the same for all galaxies , at any given time one will observe galaxies distributed uniformly over the evolutionary stage . though the plateau luminosity may be higher than the observational threshold , in the rising portion of the evolutionary curve , galaxies will be missed when the luminosity is below the threshold . in the post starburst period
, one will observe galaxies above the threshold though there is no ongoing star formation .
for example , if the luminosity is n times the observational threshold and t@xmath33 and t@xmath34 are the times when the fraction of plateau luminosity is 1/n in the rising and falling stages respectively ( see fig .
1 ) , the fraction of galaxies that will be missed is t@xmath33/t@xmath25 and the additional contribution from the post sb period is ( t@xmath34-t@xmath25)/@xmath26 . the effect depends on the sf indicator used , the duration of the sb and the luminosity function in the chosen band .
the fir emission is a good tracer of instantaneous sfr since both the rise and fall are steep . for the radio continuum
, there is a net over - estimation since the decline in the post sb period is slower . using the 1.4 ghz radio luminosity function of haarsma et al ( 2000 ) for convolution and data of fig.1
, we find the net over - estimation of the luminosity density to be 26 % .
the over - estimation in the case of uv photons is much higher .
the corrections also depend on the duration of the sb . for low values of sb duration
, the plateau region is a small fraction of the total and hence the correction is more .
thus , the uncertainties are significant as long as the sb duration is @xmath14 150 - 200 myr . in figure 2
we have plotted , the luminosity to sfr ratio normalized to the value for the imf index @xmath24 = 2.5 vs @xmath24 for three epochs , 30 myr , 95 myr and 135 myr .
it is seen that the variation with index is not the same for the different bands .
the fir luminosity shows the maximum variation , a factor of more than 5 as the index changes from 2 to 3 .
the escaping uv photons show the least change .
further , the changes also depend on the observational epoch .
thus , for individual galaxies , one can have large uncertainties in sfr if the imf varies .
if a large sample of galaxies is available , the average over varying index will result in reduced uncertainty .
it is not known whether the imf extends from high mass to low mass stars in the same region at the same time . in intense starbursts
it is believed that the imf is biased towards high mass stars . in the galactic star forming region w3 ,
megeath et al ( 1996 ) find that on extrapolation of the observed imf for m @xmath14 10 m@xmath9 , they expect to see only one star of mass @xmath8 10 m@xmath9 , but radio continuum observations show the presence of six high mass stars .
if in starbursts , the imf does not extend below , say 10 m@xmath9 , the luminosity computed would not change much , but the total stellar mass would be lower . if imf bias is present and varies from galaxy to galaxy , there will be additional uncertainties in the derived sfr .
we have computed the time evolution of luminosity in different bands of star formation indicators for the case of a continuous and constant star formation .
the luminosity rises from the time the starburst begins , reaches a plateau value and declines after the termination of the starburst .
the post - burst decline is generally , slower than the growth at early times .
we have adopted the concept of residence time to calculate the fir and escaping uv luminosities .
the fir luminosity has the sharpest profile while the uv luminosity has the broadest .
the star formation rate derived from luminosity density is subject to uncertainties depending on the shape of the imf , the residence time parameter , the duration and distribution of starburst and the observational epoch . | we have computed the time evolution of bolometric , far - infrared , h@xmath0 line , ultraviolet ( both intrinsic and escaping the star forming region ) and the nonthermal radio continuum luminosities for continuous and constant star formation terminating at 95 myr .
the luminosity rises to a plateau value and declines after the termination of starburst , but only gradually .
the time evolution profiles are broad and different for different star formation indicators .
the broad profiles lead to uncertainties in the star formation rate derived depending on the initial mass function , duration of starburst , its distribution and the observational epoch . |
exotic hadrons which include multiquark configurations provide us with important information in hadron physics .
there is a key to understand one of the most important problems in hadron physics ; what are the constituent particles of hadrons , and what are the interactions among the constituent particles at relevant low energies .
those questions are related to the fundamental questions of the qcd , such as to color confinement , chiral symmetry breaking and so on . nowadays
the exotic hadrons are studied , not only in light flavor sectors , but also in heavy flavor sectors with charm and bottom quarks @xcite .
recent experimental observations of heavy exotic hadrons , @xmath10 , @xmath11 , @xmath12 , @xmath13 in charm sector , and @xmath14 , @xmath15 in bottom sector have motivated intensive discussions about possible new dynamics in the heavy hadrons . many of those hadrons have unusual mass , decay width and branching ratios , which may not be explained as normal hadrons , such as @xmath16 and @xmath17 . as candidates of flavor exotics for future experiments ,
a new hadron state @xmath18 whose quark content is @xmath19 has been discussed theoretically @xcite .
@xmath18 is a system containing two heavy quarks and it is genuinely a flavor exotic which can not be assigned by a normal hadron . in the present paper ,
we discuss the energy spectrum of the possible bound and/or resonant states of @xmath18 . in phenomenological studies
, there exist two approaches to @xmath18 state . in one approach ,
@xmath18 is considered as a tetraquark state , in which the effective degrees of freedom are constituent quarks @xcite .
it is shown that @xmath18 may be a stable object due to the strong attraction in @xmath20 which may form a stable scalar diquark @xcite . as a result
, @xmath18 may be a deeply bound state , which does not decay through the strong interaction @xcite .
if the diquark developed , the study of @xmath18 is also useful to understand the color superconductivity in high density quark matter @xcite .
the tetraquark states such as @xmath21 , @xmath22 and @xmath23 states with @xmath24 as well as @xmath25 with @xmath26 have been discussed also as stable objects @xcite .
another approach is the hadronic molecule picture .
when four quarks ( @xmath19 ) are present , they may form hadronic clusters ( @xmath27 ) which may alternatively become relevant degrees of freedom instead of diquarks @xcite .
the hadronic molecule picture is applied to the energy region close to the thresholds . in the @xmath28 system
, two mesons composed by @xmath27 , the pseudoscalar meson @xmath29 and the vector meson @xmath30 , can become effective degrees of freedom as constituents . in the heavy quark limit , the pseudoscalar meson @xmath31 and
the vector meson @xmath32 become degenerate in mass , and hence both of them should be considered .
hereafter we introduce the notation @xmath33 for @xmath31 and @xmath32 .
indeed , we will show that the mass degeneracy of the @xmath31 and @xmath32 activates the one - pion exchange potential between two @xmath33 s , and the bound and/or resonant @xmath34 states are formed .
the tetraquark picture and the hadronic molecule picture are quite different .
the tetraquark picture is applied to the deeply bound state . to apply this picture to the shallow bound states or resonant states , slightly below or above thresholds
respectively , is not appropriate , because the continuous @xmath33 states above thresholds are not taken into account . on the other hand ,
the hadronic molecule picture can be applied to the shallow and resonant states , while it can not be applied to the deeply bound states . though there have been several studies for bound states of @xmath34 in the hadronic molecule picture ,
resonant states of @xmath34 have not been studied yet .
moreover , the quantum numbers @xmath35 which have been discussed are limited only @xmath36 . to be more problematic ,
several channels in coupled channel problem have been neglected .
the last point is very important in the heavy quark systems . in the heavy quark limit
, we need to consider the mass degeneracy of @xmath31 and @xmath32 which provides more number of channels than discussed in the literature . in the present work , we discuss both bound and resonant states of @xmath34 systematically for various quantum numbers @xmath35 up to @xmath37 by considering the fully coupled channel problem .
the systematic analysis of the energy spectrum for various quantum numbers is important to investigate the dynamics governing the systems .
indeed , it will be shown that there are several new shallow bound and/or resonant states , which were not found in other studies , thanks to the strong attraction induced from the channel couplings even for larger @xmath38 or angular momentum .
this paper is organized as follows . in section 2 ,
we give the interaction between two @xmath33 mesons with respecting the heavy quark symmetry and chiral symmetry .
we introduce the two types of potentials , the @xmath39 exchange potential and @xmath40 exchange potential . in section 3 ,
we classify all the @xmath34 systems up to @xmath41 , and search the bound and/or resonant states by applying the potentials and solving the schrdinger equations numerically . in section 4
, we compare our results from the hadronic molecule picture with the previous results from the tetraquark picture .
we summarize our discussions in the final section .
the dynamics of the hadronic molecule of @xmath34 respects two important symmetries ; the heavy quark symmetry and chiral symmetry .
the heavy quark symmetry induces the mass degeneracy of @xmath31 and @xmath32 in the heavy quark limit .
because of this , we have to consider the channels of degenerate pairs , such as @xmath42 , @xmath43 , @xmath44 and @xmath45 , leading to the mixing of them ; @xmath43-@xmath44 , @xmath45-@xmath45 , @xmath42-@xmath45 , @xmath43-@xmath45 . as for the meson - exchange interaction between two @xmath33 s , the one pion exchange potential ( opep ) exists at long distances .
the existence of a pion is a robust consequence of spontaneous breaking of chiral symmetry @xcite .
the opep is provided by the @xmath46 and @xmath47 vertices whose coupling strengths are equally weighted thanks to the heavy quark symmetry .
we note that there is no @xmath48 vertex because of the parity conservation .
we should keep in mind in the following discussions that the existence of both @xmath46 and @xmath47 vertices thanks to the heavy quark symmetry provide the channel mixings in @xmath42 , @xmath43 , @xmath44 and @xmath45 at long distance . at short distances , heavier mesons exchange potential , which provide similar channel mixings ,
should also be considered . to derive the @xmath34 potential
, we employ the effective lagrangians based on the heavy quark symmetry and chiral symmetry @xcite .
they describe the interaction between heavy mesons which are given by the exchange of pion and vector mesons ( @xmath49 ) .
the interaction lagrangians are given as @xmath50 where @xmath51 , @xmath52 and @xmath53 , @xmath54 .
the subscripts @xmath55 and @xmath56 are for light flavor indices , up and down , and @xmath57 is a four - velocity which will be fixed as @xmath58 below .
the pion and vector meson ( @xmath59 and @xmath60 ) fields are defined by @xmath61 following ref .
@xcite , the coupling constants in the interaction lagrangians are given as @xmath62 where @xmath63 is the pion decay constant and @xmath64 is the mass of the @xmath59 meson .
the opeps are derived by the interaction lagrangians ( [ eq : pipp * ] ) and ( [ eq : pip*p * ] ) as follows : @xmath65 \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : pot_pp*p*p } \\ v^{\pi}_{p_{1}^{\ast}p_{2}^{\ast } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & \left ( \sqrt{2 } \frac{g}{f_{\pi } } \right)^{2 } \frac{1}{3 } \left [ \vec{t}_{1 } \!\cdot\ !
\vec{t}_{2 } \ , c(r;m_{\pi } ) \!+\ !
s_{t_{1},t_{2 } } \ , t(r;m_{\pi } ) \right ] \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : pot_p*p*p*p * } \\ v^{\pi}_{p_{1}p_{2 } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & \left ( \sqrt{2 } \frac{g}{f_{\pi } } \right)^{2 } \frac{1}{3 } \left [ \vec{\varepsilon}_{1}^{\,\ast } \!\cdot\ !
\vec{\varepsilon}_{2}^{\ , \ast } \ , c(r;m_{\pi } ) \!+\ !
s_{\varepsilon_{1}^{\ast},\varepsilon_{2}^{\ast } } \ , t(r;m_{\pi } ) \right ] \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : pot_ppp*p * } \\ v^{\pi}_{p_{1}p_{2}^{\ast } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & -\left ( \sqrt{2 } \frac{g}{f_{\pi } } \right)^{2 } \frac{1}{3 } \left [ \vec{\varepsilon}_{1}^{\,\ast } \!\cdot\ !
\vec{t}_{2 } \ , c(r;m_{\pi } ) \!+\ !
s_{\varepsilon_{1}^{\ast},t_{2 } } \ , t(r;m_{\pi } ) \right ] \vec{\tau}_{1 } \!\cdot\ ! \vec{\tau}_{2 } ,
\label{eq : pot_pp*p*p*}\end{aligned}\ ] ] where @xmath66 is the pion mass . here
three polarizations are introduced for @xmath67 as defined by @xmath68 and @xmath69 , and the spin - one operator @xmath70 is defined by @xmath71 . as a convention , we assign @xmath72 for an incoming vector particle and @xmath73 for an outgoing vector particle . here @xmath74 and @xmath75 are isospin operators for @xmath76 and @xmath77 ; @xmath78 and 1 for the @xmath79 and @xmath80 channels , respectively .
we define the tensor operators @xmath81 where @xmath82 is a unit vector between the two mesons .
the @xmath59 meson exchange potentials are similarly obtained from the interaction lagrangians ( [ eq : vpp])-([eq : vp*p * ] ) , @xmath83 \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : rhopot_bb*b*b } \\ v^{\rho}_{p_{1}^{\ast}p_{2}^{\ast } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & \left ( 2\lambda g_v \right)^{2 } \frac{1}{3 } \left [ 2\vec{t}_{1 } \!\cdot\ ! \vec{t}_{2 } \ , c(r;m_{\rho } ) \!-\ !
s_{t_{1},t_{2 } } \ , t(r;m_{\rho } ) \right ] \vec{\tau}_{1 } \!\cdot\
! \vec{\tau}_{2 } \nonumber \\ & & + \left ( \frac{\beta g_v}{2m_{\rho } } \right)^{2 } \frac{1}{3 } c(r;m_{\rho } ) \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : rhopot_b*b*b*b * } \\
v^{\rho}_{p_{1}p_{2 } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & \left ( 2\lambda g_v \right)^{2 } \frac{1}{3 } \left[2 \vec{\varepsilon}_{1}^{\,\ast } \!\cdot\ !
\vec{\varepsilon}_{2}^{\ , \ast } \
, c(r;m_{\rho } ) \!-\ !
s_{\varepsilon_{1}^{\ast},\varepsilon_{2}^{\ast } } \ , t(r;m_{\rho } ) \right ] \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2 } , \label{eq : rhopot_bbb*b * } \\ v^{\rho}_{p_{1}p_{2}^{\ast } \rightarrow p_{1}^{\ast}p_{2}^{\ast } } & \!=\ ! & -\left ( 2\lambda g_v \right)^{2 } \frac{1}{3 } \left [ 2\vec{\varepsilon}_{1}^{\,\ast } \!\cdot\ ! \vec{t}_{2 } \ , c(r;m_{\rho } ) \!-\ !
s_{\varepsilon_{1}^{\ast},t_{2 } } \ , t(r;m_{\rho } ) \right ] \vec{\tau}_{1 } \!\cdot\ !
\vec{\tau}_{2}. \label{eq : rhopot_bb*b*b*}\end{aligned}\ ] ] the @xmath60 meson exchange potentials are obtained by replacing the mass of @xmath59 meson with the one of @xmath60 meson and by removing the isospin factor @xmath84 .
the opep s of @xmath34 differ from the ones of @xmath85 in that the overall signs are changed due to @xmath86-parity .
the situation is the same with @xmath60 meson exchange potentials , while @xmath59 meson exchange potentials of @xmath85 are not changed because the @xmath86-parity is even @xcite . in the above equations , @xmath87 and @xmath88
are defined as @xmath89 with @xmath90 .
we introduce the monopole type form factor at each vertex to take into account of the size effect of @xmath91 mesons .
then the function reflected form factors is defined as @xmath92 where @xmath93 and @xmath94 are the mass and three - momentum of the exchanged meson @xmath95 ( @xmath96 ) and @xmath97 is the cut - off parameter .
the cut - off parameter @xmath97 are determined from the size of @xmath31 estimated from the constituent quark model as discussed in refs .
the cut - off parameters are @xmath98 mev and @xmath99 mev when the @xmath39 exchange potential is employed , while @xmath100 mev and @xmath101 mev when the @xmath40 is employed . up to now
we have given the meson - exchange potentials between two @xmath33 mesons .
we should note that the potentials contain spin operators and tensor operators , hence that the potentials for each quantum number are different . in the next section ,
we classify all the @xmath34 states up to @xmath41 and give the corresponding potentials in matrix forms .
let us classify all the possible quantum numbers of the @xmath34 systems with isospin @xmath102 , total angular momentum @xmath38 ( @xmath37 ) , and parity @xmath103 .
we need also principal quantum number @xmath104 , @xmath105 , @xmath106 , if there exist several bound states for a given @xmath35 .
we show the quantum numbers @xmath107 and the channels in the wave functions in table [ tbl : classification ] . it is noted that the wave functions must be symmetric under the exchange of the two @xmath33 mesons .
we use the notation @xmath108 to indicate the states with the internal spins @xmath109 and angular momentum @xmath110 .
for example , the @xmath111 state is a superposition of four channels ; @xmath112 , @xmath113 , @xmath114 and @xmath115 .
all the possible channels should be mixed for a given the quantum number . in the previous studies , the channel mixings were not fully considered @xcite . here
we pay an attention to that the approximate mass degeneracy of @xmath31 and @xmath32 plays a crucial role to mix the channels .
otherwise the attraction from the mixing effect becomes suppressed .
we note that the tensor force is also important to mix the channels with different angular momenta , @xmath110 and @xmath116 . as a result
, we obtain the hamiltonian in a matrix form with the basis of those coupled channels .
the explicit forms of the hamiltonian for each @xmath35 are summarized in appendix a. now we are ready to solve the coupled channel schrdinger equations for each quantum number .
the renormalized numerov method @xcite is adopted to numerically solve the coupled second - order differential equations .
the resonant states are identified by the behavior of the phase shift @xmath117 as a function of the scattering energy @xmath118 .
the resonance position @xmath119 is defined by an inflection point of the phase shift @xmath120 and the resonance width by @xmath121 following refs .
@xcite . to check the consistency of our numerical calculations
, we also adopt the complex scaling method ( csm ) , in which the resonant state is defined as a pole in the complex energy plane @xcite .
we obtain an agreement in the results of the renormalized nemerov method and the csm .
we summarize our numerical results for @xmath122 bound / resonant states in table [ tbl : result_table_dd ] and fig [ fig : resultdd ] . in @xmath122 states ,
we find several bound and/or resonant states in @xmath79 , while there is no bound state in @xmath80 .
in general , the attractive force of pion exchange in @xmath80 is three times weaker than in @xmath79 due to the isospin factor . as a numerical result , @xmath122 bound states in @xmath80 are not obtained but only resonant states are .
let us look at our results one by one for each quantum number in detail . in the following text ,
most of the numerical values are those for the case of the @xmath123 potential , because the results from the @xmath123 potential are generally not so different from those from the @xmath39 potential , except for the @xmath124 state .
the energies are measured from the threshold , which is defined to be the lowest mass among the channels for a given quantum number as tabulated in table [ tbl : classification ] .
for example , we adopt the @xmath125 mass as threshold for @xmath126 , while , the @xmath2 mass for @xmath127 . *
this state has only one channel of @xmath125 ( see table .[tbl : classification ] ) and the pion exchange potential is attractive as shown in eq .
( [ eq : pi_pot_00- ] ) . as a result
, the very deep bound state of @xmath125 is generated with binding energy 132.1 mev measured from the @xmath125 threshold .
* the pion exchange potential is repulsive for diagonal components as shown in eq .
( [ eq : pi_pot_1 + - ] ) .
however this state has four components and the mixing of the @xmath109- and @xmath128-waves causes the strong tensor attraction from the off - diagonal components of the potential .
consequently , there is a deeply bound state of mostly @xmath125 with binding energy 62.3 mev measured from the @xmath125 threshold .
* there are twin shape resonances of @xmath2 with the resonance energy 17.8 mev and the decay width 41.6 mev for the first resonance , and the resonance energy 152.8 mev and the decay width 10.6 mev for the second .
the resonance energies are measured from the @xmath2 threshold .
those resonances are formed by the centrifugal barrier in the @xmath103-wave . *
this state contains only @xmath128-wave components of @xmath125 and @xmath129 .
the potential is weakly attractive .
nevertheless , due to the centrifugal barrier in the @xmath128-wave , there is a shape resonance of @xmath125 scattering at the energy 33.7 mev from the @xmath125 threshold , but the decay width 196.3 mev is very wide .
* when the opep is employed , there are twin resonant states with the resonance energy 0.1 mev and the decay width 0.02 mev for the first resonance , and the resonance energy 118.0 mev and the decay width 23.4 mev for the second from the @xmath125 threshold .
when the effects of @xmath59 and @xmath60 meson exchange are included , the first resonance becomes a weakly bound state with the binding energy 4.3 mev , because the @xmath59 meson exchange enhances the central force attraction of the pion exchange .
the @xmath60 meson exchange plays a minor role due to the isospin factor , although this contribution suppresses the attractive central force .
the second resonant state with the resonance energy 112.1 mev and the decay width 26.6 mev is not affected very much . from the analysis of wave function components of the two bound states
, we have verified that the lower and higher states are dominated by @xmath125 and @xmath129 , respectively .
* this is the only @xmath80 state in @xmath130 .
the interaction in @xmath80 are either repulsive or only weakly attractive as already discussed . nevertheless , due to the @xmath103-wave centrifugal barrier
, we find twin shape resonances ; the resonance energy 2.3 mev and the decay width 37.4 mev for the first resonance , and the resonance energy 144.2 mev and the decay width 34.4 mev for the second , from the @xmath125 threshold .
here several comments are in order .
first , we have obtained several bound and/or resonant state even for @xmath41 . here
the long range force by the opep becomes effective for the extended objects with large angular momenta .
it is also interesting to have twin states " for several quantum numbers , @xmath131 , @xmath124 and @xmath132 .
we have to note that the channel couplings by @xmath133 and @xmath134 are important to produce the obtained energy spectrum . indeed , if we cut the channel coupling , we confirm that many of the states disappear .
thus , we consider that the pattern of the energy spectrum is reflected by the dynamics of the fully coupled channels .
second , the present formalism of the hadronic molecule picture can not be applied to compact objects .
when the two @xmath33 mesons overlap spatially , we need to consider the internal structure of @xmath33 , which is not included in the present hadronic picture .
therefore we shall adopt 1 fm or larger for the size of the bound state to be interpreted as a molecular state .
the size of 1 fm is twice of typical radius of @xmath135 fm .
for instance , the @xmath136 bound state with the binding energy @xmath137 mev is identified with a molecular state because it has the size @xmath138 fm , while the @xmath139 bound state with @xmath140 mev is not because its size is @xmath141 fm .
we emphasize , however , that this criterion is not definitive but gives only qualitative guide .
next we discuss the @xmath142 states .
we use the same coupling constants , and change only the masses of heavy mesons with small difference of the cutoff parameters .
the results are summarized in table [ tbl : result_table_bb ] and figs .
[ fig : resultbb0 ] and [ fig : resultbb1 ] . at first glance , we find that the @xmath142 states have many bound and resonant states in comparison with the @xmath122 states .
there are two reasons .
first , the kinetic term is suppressed in the hamiltonian because the reduced mass becomes larger in the bottom sector .
second , the effect of channel - couplings becomes more important , because a pseudoscalar meson @xmath143 and a vector meson @xmath144 become more degenerate thanks to the heavy quark symmetry ; similar discussion has been done in refs .
we note that , as a consequence of the strong attraction , several new states appear in the @xmath142 states in isospin triplet ; @xmath145 , @xmath146 , @xmath147 and @xmath148 .
the corresponding states are not obtained in the @xmath122 states .
as noted in the charm sector , the deeply bound states will not be within the scope of the hadronic molecule picture .
in the bottom sector , due to more attraction , more bound states are generated .
for example , we have three states ( @xmath104 , @xmath105 , @xmath149 ) in the @xmath150 state .
however , the @xmath104 , @xmath105 states are too compact ( 0.5 fm and 0.9 fm , respectively ) to be considered as hadronic molecules .
the @xmath151 state is an extended object ( 2.5 fm ) , and hence can be considered as the hadronic molecule .
similarly , it would not be conclusive yet to consider hadronic molecules for the following states with radii less than 1 fm ; the @xmath104 state in @xmath152 , the @xmath104 , 1 states in @xmath131 , the @xmath104 state in @xmath153 , the @xmath104 , 1 states in @xmath124 and the @xmath104 state in @xmath154 .
.possible channels of @xmath155 for a set of quantum numbers @xmath102 and @xmath156 for @xmath37 .
[ cols="^,^,^",options="header " , ] [ tbl : result_table_bb ] bound and resonant states for various @xmath35 .
solid lines are for our predictions with numerical values as denoted above the lines , and the values in parentheses below the lines denote the decay width @xmath157 of the resonances when the @xmath158 potential is employed .
mass values are given in units of mev .
compact objects which can not be regarded as molecular states are shown with grey lines.,width=491 ] bound and resonant states around the thresholds with @xmath35 in @xmath79 .
compact objects which can not be regarded as molecular states are shown below the wavy line , but their location do not reflect correct energy scale .
( same convention as fig .
[ fig : resultdd].),width=491 ] bound and resonant states around the thresholds with @xmath35 in @xmath80 .
( same convention as fig .
[ fig : resultdd].),width=491 ]
in the present paper , we have considered @xmath28 as a hadronic molecule composed by @xmath34 in which one pion exchange potential induces a dominant attraction . on the other hand ,
as mentioned in the introduction , @xmath28 can also be considered as a tetraquark @xmath19 in which a diquark @xmath20 provides a strong binding energy .
the different features between the hadronic molecule and tetraquark pictures are seen in their sizes . for hadronic molecules to be valid ,
hadron constituents are sufficiently far apart such that their identities as hadrons must be maintained
. they can not be too close to overlap each other .
therefore , masses of hadronic molecules should appear around their threshold regions .
in contrast , tetraquarks may be strongly bound and become compact objects as genuine quark objects .
thus , their natures are differentiated in the masses , small binding energy of order ten mev or less , or larger one . although it is tempting to seek for a framework to cover both scales , such a problem is out of scope of the present paper .
instead , we compare the results from the two pictures and just clarify the differences between them . in the hadronic molecule picture by @xmath34 ,
as presented in the previous section , we obtain , not only the bound states , but also the resonant states in many @xmath35 quantum numbers . both in charm and bottom sectors
, it is remarkable that there are even the twin states in several quantum numbers , @xmath131 , @xmath124 and @xmath132 . in the tetraquark picture including a diquark model @xcite , in contrast , only two bound states in @xmath159 and @xmath148 have been predicted until now , as discussed for an example in refs .
@xcite as recent works . for @xmath152
, the predicted binding energy can be around 70 mev from the @xmath125 threshold . for @xmath148
, the predicted binding energy can be around 27 mev .
the mass of the @xmath152 states in the tetraquark picture is accidentally close to our value 62 mev in the hadronic molecule picture .
however , this comparison should be considered more carefully . because the size of the @xmath152 state is 0.8 fm from table [ tbl : result_table_dd ] , the @xmath133 and @xmath134 mesons composing the @xmath152 state
would be overlapped spatially if the size of each @xmath133 and @xmath134 meson is a scale of about 1 fm .
in such a compact object , the quark degrees of freedom may become active to contribute to the dynamics like the tetraquark picture .
however , such an effect is out of the scope of the present hadronic molecule picture .
the @xmath148 state can not be found in our present study . in any cases , the feature in the energy spectrum in the @xmath130 molecule picture is that there are many shallow bound states and resonances in @xmath131 , @xmath153 , @xmath124 and @xmath132 , which are not found in the tetraquark picture .
thus , we observe that the energy spectrum of the hadronic molecule picture by @xmath34 is qualitatively different from that of the tetraquark picture .
we have discussed exotic mesons with double charm and bottom flavor whose quark content @xmath160 is genuinely exotic .
we have taken the hadronic picture , and considered molecular states of two heavy mesons @xmath91 s ( a pseudoscalar meson @xmath161 , @xmath143 , and a vector meson @xmath162 , @xmath163 ) .
with respecting the heavy quark and chiral symmetries , we have constructed the @xmath39 exchange potential and the @xmath123 exchange potential between the two heavy mesons . to investigate the bound and/or resonant states , we have numerically solved the coupled channel schrdinger equations for the @xmath34 states with @xmath8 for @xmath164 . as results ,
we have found many bound and/or resonant states in both charm and bottom sectors . the @xmath122 bound and resonant states have moderate energies and decay widths around the thresholds in several channels with quantum numbers ; @xmath150 , @xmath152 , @xmath131 , @xmath153 , @xmath124 and @xmath132 .
the @xmath142 states have more bound and resonant states with various quantum numbers .
several new states appear in the @xmath142 states in isotriplet states , such as @xmath154 , @xmath146 , @xmath147 and @xmath148 , which can not be found in the charm sector .
by contrast to the @xmath122 states , some @xmath142 states are very compact objects with a large binding energy much below the thresholds . perhaps , these states can not survive as hadronic molecules and more consideration of quark dynamics such as tetraquarks is required . the energy spectrum for quantum numbers
@xmath35 will help us to study the structure of the exotic states with @xmath160 . in the present hadronic molecule picture
, many shallow bound states and resonant states appear around the thresholds in several quantum numbers .
indeed , they were not found in the tetraquark picture .
it is interesting to note that , around the thresholds for @xmath131 , @xmath153 , @xmath124 and @xmath132 , the shape of energy spectrum in those states in the charm sector seems to that in he bottom sector .
it may indicate a universal behavior of the energy spectrum around the thresholds .
experimental studies of those exotic hadrons should be performed in the coming future .
the double charm production in accelerator facilities will help us to search them @xcite .
recently it has been discussed that the quark - gluon plasma in the relativistic heavy ion collisions could produce much abundance of exotic hadrons including the exotic mesons with double charm @xcite .
those experimental studies will shed a light on the nature of the exotic mesons with double charm and bottom flavor , and provide important hints to the fundamental questions of the strong interaction in hadron physics .
we thank prof . s. takeuchi and prof .
m. takizawa for fruitful discussions and comments .
this work is supported in part by grant - in - aid for scientific research on priority areas elucidation of new hadrons with a variety of flavors ( e01 : 21105006 ) " ( s.y . and a.h . ) and by grant - in - aid for young scientists ( b ) 22740174 " ( k.s .
) , from the ministry of education , culture , sports , science and technology of japan .
the hamiltonian is a sum of the kinetic term and potential term as , @xmath165 for the @xmath39 exchange potential only , and @xmath166 for the @xmath158 potential . the kinetic terms with including the explicit breaking of the heavy quark symmetry by the mass difference @xmath167 are @xmath168 where @xmath169 with integer @xmath170 , @xmath171 , @xmath172 , @xmath173 .
e. s. swanson , phys .
rept . * 429 * , 243 - 305 ( 2006 ) .
m. b. voloshin , prog . part .
* 61 * , 455 - 511 ( 2008 ) .
m. nielsen , f. s. navarra , s. h. lee , phys .
* 497 * , 41 - 83 ( 2010 ) .
n. brambilla _ et al .
. phys .
j. c * 71 * , 1534 ( 2011 ) [ arxiv:1010.5827 [ hep - ph ] ] .
s. zouzou , b. silvestre - brac , c. gignoux , j. m. richard , z. phys .
* c30 * , 457 ( 1986 ) .
s. pepin , f. stancu , m. genovese and j. m. richard , phys .
b * 393 * , 119 ( 1997 ) [ arxiv : hep - ph/9609348 ] .
d. m. brink and f. stancu , phys .
d * 57 * , 6778 ( 1998 ) .
b. silvestre - brac and c. semay , z. phys .
c * 59 * , 457 ( 1993 ) .
s. pepin , f. stancu , m. genovese and j. m. richard , phys .
b * 393 * , 119 ( 1997 ) [ arxiv : hep - ph/9609348 ] .
j. schaffner - bielich and a. p. vischer , phys .
d * 57 * ( 1998 ) 4142 [ arxiv : nucl - th/9710064 ] .
j. vijande , e. weissman , n. barnea and a. valcarce , phys .
d * 76 * , 094022 ( 2007 ) [ arxiv:0708.3285 [ hep - ph ] ] .
j. vijande , e. weissman , a. valcarce , n. barnea , phys .
rev . * d76 * , 094027 ( 2007 ) .
[ arxiv:0710.2516 [ hep - ph ] ] .
d. ebert , r. n. faustov , v. o. galkin and w. lucha , phys .
d * 76 * , 114015 ( 2007 ) [ arxiv:0706.3853 [ hep - ph ] ] .
f. s. navarra , m. nielsen and s. h. lee , phys .
b * 649 * , 166 ( 2007 ) [ hep - ph/0703071 ] .
s. h. lee , s. yasui , w. liu , c. m. ko , eur .
j. * c54 * , 259 - 265 ( 2008 ) .
[ arxiv:0707.1747 [ hep - ph ] ] .
m. zhang , h. x. zhang and z. y. zhang , commun .
phys . *
50 * , 437 ( 2008 ) [ arxiv:0711.1029 [ nucl - th ] ] .
y. yang , c. deng , j. ping and t. goldman , phys . rev .
d * 80 * , 114023 ( 2009 ) .
j. vijande , a. valcarce , n. barnea , phys .
rev . * d79 * , 074010 ( 2009 ) .
[ arxiv:0903.2949 [ hep - ph ] ] .
j. vijande , a. valcarce and t. f. carames , few body syst . *
50 * , 195 ( 2011 ) .
j. p. ader , j. m. richard and p. taxil , phys .
d * 25 * , 2370 ( 1982 ) .
j. vijande , a. valcarce and j. m. richard , phys .
d * 76 * , 114013 ( 2007 ) [ arxiv:0707.3996 [ hep - ph ] ] .
a. v. manohar and m. b. wise , nucl .
b * 399 * ( 1993 ) 17 [ arxiv : hep - ph/9212236 ] .
n. a. tornqvist , z. phys .
c * 61 * , 525 ( 1994 ) [ hep - ph/9310247 ] .
. ding , j. -f .
liu and m. -l .
yan , phys .
d * 79 * , 054005 ( 2009 ) [ arxiv:0901.0426 [ hep - ph ] ] .
r. molina , t. branz and e. oset , phys .
d * 82 * , 014010 ( 2010 ) [ arxiv:1005.0335 [ hep - ph ] ] .
r. l. jaffe , phys .
* d15 * , 267 ( 1977 ) .
g. burdman and j. f. donoghue , phys .
b * 280 * , 287 ( 1992 ) .
m. b. wise , phys .
d * 45 * , 2188 ( 1992 ) .
t. m. yan , h. y. cheng , c. y. cheung , g. l. lin , y. c. lin and h. l. yu , phys .
d * 46 * , 1148 ( 1992 ) [ erratum - ibid .
d * 55 * , 5851 ( 1997 ) ] .
r. casalbuoni , a. deandrea , n. di bartolomeo , r. gatto , f. feruglio and g. nardulli , phys .
* 281 * , 145 ( 1997 ) [ arxiv : hep - ph/9605342 ] .
a. v. manohar and m. b. wise , camb .
monogr . part .
nucl . phys .
cosmol .
* 10 * ( 2000 ) 1 . c. isola , m. ladisa , g. nardulli and p. santorelli , phys
. rev .
d * 68 * , 114001 ( 2003 ) [ arxiv : hep - ph/0307367 ] . s. yasui and k. sudoh , phys .
d * 80 * , 034008 ( 2009 ) [ arxiv:0906.1452 [ hep - ph ] ] .
y. yamaguchi , s. ohkoda , s. yasui and a. hosaka , phys .
d * 84 * , 014032 ( 2011 ) [ arxiv:1105.0734 [ hep - ph ] ] .
s. ohkoda , y. yamaguchi , s. yasui , k. sudoh and a. hosaka , arxiv:1111.2921 [ hep - ph ] .
y. yamaguchi , s. ohkoda , s. yasui and a. hosaka , arxiv:1111.2691 [ hep - ph ] .
b.r.johnson , j. chem .
69(10 ) , 15 nov .
1978 k. arai and a. t. kruppa , phys .
c * 60 * , 064315 ( 1999 ) .
a. del fabbro , d. janc , m. rosina and d. treleani , phys .
d * 71 * , 014008 ( 2005 ) [ hep - ph/0408258 ] .
h. nagahiro , k. nawa , s. ozaki , d. jido and a. hosaka , phys .
d * 83 * , 111504 ( 2011 ) [ arxiv:1101.3623 [ hep - ph ] ] . | we study exotic mesons with double charm and bottom flavor ( @xmath0 ) , whose quark configuration is @xmath1 .
this quark configuration has no annihilation process of quark and antiquark , and hence is a genuinely exotic states .
we take a hadronic picture by considering the molecular states composed of a pair of heavy mesons , such as @xmath2 , @xmath3 and @xmath4 for charm flavor , and @xmath5 , @xmath6 and @xmath7 for bottom flavor .
the interactions between heavy mesons are derived from the heavy quark effective theory .
all molecular states are classified by @xmath8 quantum numbers , and are systematically studied up to the total angular momentum @xmath9 . by solving the coupled channel schrdinger equations , due to the strong tensor force of one pion exchanging ,
we find bound and/or resonant states of various quantum numbers . |
a system of atoms interacting with a common reservoir ( the electromagnetic vacuum ) is often treated using an assumption of independent dissipation rates for the atoms .
this is a valid assumption for the case when the interatomic distances are large ( compared to an optical wavelength ) ; however , when the distances between atoms become smaller or comparable to an optical wavelength , the mode structure around one atom is changed due to the presence of other atoms located in its immediate vicinity , and the decay rates are modified .
equivalently , the radiation emitted by one atom can be scattered off the second atom , thus changing the radiative properties of the system . as a consequence
, the radiative decay of the ensemble must be viewed as a cooperative effect .
a quantitative analysis of cooperative effects was given by dicke @xcite for an ensemble of two - level systems confined to a spherical volume whose radius is much smaller than an optical wavelength .
for such a system , the collective decay rate can be increased by a factor proportional to the number of atoms in the ensemble .
further investigations extend the treatment to arbitrary interatomic distances , although the calculations become more complex .
formalisms for the treatment of cooperative spontaneous emission and resonance fluorescence from a system of many atoms have been developed @xcite , and have been applied to a system of two two - level atoms @xcite .
cooperative decay in multilevel atomic systems has yet to be treated in detail . a multilevel atom has properties not possessed by a two - level atom : for example , it can store information in superpositions of ground state magnetic sublevels .
such ensembles are extensively discussed in the literature as convenient systems for information storage or large scale entanglement generation @xcite . in particular ,
pencil - shaped media have been used for the generation of spin squeezed and schrodinger cat states , in the context of continuous measurement of a scattered field @xcite . in such schemes , the coupling to the vacuum has a two - fold function : on one hand it gives rise to the signal while , on the other hand , it leads to an irreversible leakage of information from the system to the environment . in treating the losses due to spontaneous emission , the above mentioned assumption of independent atoms
is generally used , which is a sound assumption as long as the atomic density is low . however
, optimal results ( e.g. strong entanglement ) are found in the regime of resonant optical depths greater than unity , a regime in which pencil - shaped media of two - level atoms exhibit superradiant behavior @xcite .
even if the dynamics of a collection of multilevel atoms might be substantially different from that of the two - level ensemble , the validity of the independent spontaneous emission regime is questionable .
we proceed in the present publication with an analysis of cooperative effects in a system of two , four - level atoms .
this provides the starting point for an extension to many atom systems , while it also addresses the non - trivial question of the importance of cooperative decoherence in a simple quantum information system of two qubits .
the calculations are performed for a @xmath0 transition irradiated with a monochromatic , off - resonant , @xmath1 polarized , classical laser field .
the collective decoherence of an initial equal superposition of ground sublevels ( @xmath2 polarized atomic state ) is obtained for arbitrary interatomic separations and compared with that of atoms independently coupled to the reservoir .
in addition , the transfer of population from a @xmath3 polarized atomic state with both atoms in one of the ground sublevels to another @xmath3 polarized state with both atoms in the other sublevel , is analyzed .
a polarization swap effect is also discussed , where an @xmath2 polarized atom induces @xmath2 coherence in a neighboring atom @xcite .
the paper is organized as follows : in sec .
ii the theoretical method used is described . in sec .
iii analytical solutions for the case of close atoms are obtained .
numerical solutions for arbitrary separations are discussed and plotted in sec .
iv . in sec . v the polarization swap effect for arbitrary separations is discussed , while sec .
vi contains some conclusions .
as indicated in fig . [ scheme](a ) , the two atoms ( having natural frequency @xmath4 ) are located at the origin , @xmath5 , and at @xmath6 , respectively . the traveling wave driving field propagates in the positive @xmath3 direction with wave vector @xmath7 , frequency @xmath8 ( detuned from @xmath9by @xmath10 ) , and circular polarization @xmath1 .
[ h ] fig1.eps denoting the @xmath11 states of a single atom with @xmath12 ( ground , @xmath13 eigenvalue ) , @xmath14 ( ground , @xmath15 eigenvalue ) , @xmath16 ( excited , @xmath17 eigenvalue ) and @xmath18 ( excited , @xmath19 eigenvalue ) , the classical field drives the @xmath20 @xmath18 transitions .
the hamiltonian for the system ( composed of the two atoms indexed with @xmath21 , where @xmath22 ) is a sum of the free hamiltonian ( @xmath23 ) , the classical field - atom interaction hamiltonian ( @xmath24 ) and quantized bath - atom interaction hamiltonian ( @xmath25 ) given by:@xmath26 + \underset{\mathbf{k},\lambda}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \hbar\omega_{k}a_{\mathbf{k},\lambda}^{\dagger}a_{\mathbf{k},\lambda~},\\ v_{c~ } & = \underset{i=1,2}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \hbar\chi\left [ \left\vert \alpha\right\rangle _ { i}\left\langle \downarrow\right\vert _ { i~}e^{i\mathbf{k}_{l}\mathbf{\cdot r}_{i~}}e^{-i\omega t}+\left\vert \downarrow\right\rangle _ { i}\left\langle \alpha\right\vert _ { i~}e^{-i\mathbf{k}_{l}\mathbf{\cdot r}_{i~}}e^{i\omega t}\right ] , \nonumber\\ v_{q~ } & = \underset{i=1,2}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \left [ \underset{j=\downarrow,\uparrow}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \underset{j^{\prime}=\beta,\alpha}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \underset{\mathbf{k},\lambda}{% % tcimacro{\tsum } % % beginexpansion { \textstyle\sum } % endexpansion } \hbar\left
[ g_{\mathbf{k},\lambda~}^{jj^{\prime}}\left\vert j\right\rangle _ { i}\left\langle j^{\prime}\right\vert _ { i~}a_{\mathbf{k},\lambda~}^{\dagger } e^{-i\mathbf{k\cdot r}_{i~}}+\left ( g_{\mathbf{k},\lambda~}^{jj^{\prime}% } \right ) ^{\ast}\left\vert j^{\prime}\right\rangle _ { i}\left\langle j\right\vert _ { i~}a_{\mathbf{k},\lambda~}e^{i\mathbf{k\cdot r}_{i~}}\right ] \right ] .\nonumber\end{aligned}\ ] ] as usual , the field is quantized in a volume @xmath27 and is described by the creation and annihilation operators @xmath28and @xmath29 , corresponding to modes with wave vector @xmath7 and polarization @xmath30 .
the atomic dipole moment @xmath31 ( the same for both atoms ) couples to both the classical and quantized fields .
the cw driving field at the position of the atoms ( @xmath32 for @xmath22 ) is expressed as @xmath33 \widehat{\mathbf{\epsilon}}_{+~}$ ] .
the classical part of the interaction contains the rabi frequency defined as:@xmath34 where @xmath35 is the matrix element @xmath36 ; the interaction with the quantum vacuum has an associated coupling strength@xmath37 proportional to the dipole matrix element in the direction of the unit polarization vector @xmath38 .
the quantities that are relevant in what follows are the collective coherence operator@xmath39 and the normalized population operator@xmath40 the derivation of the time evolution of the expectation values of these two operators is the goal of our calculations . they can expressed in terms of density matrix elements as@xmath41
@xmath42 we consider two problems : first , one with both atoms prepared initially in a superposition of ground states with maximum coherence @xmath43 ( collective coherence equal to @xmath44 ) and , second , one where the population is transferred from the state with both atoms in the @xmath45 state to the one with both atoms in the @xmath14 state .
qualitative ( and some quantitative ) details of the calculations are described in the following .
the subspace of interest in which the collective operators defined above act ( henceforth named the ground subspace ) is of dimension @xmath11 and it is spanned by state vectors containing the ground substates of the two atoms ( @xmath46,@xmath47 , @xmath48 and @xmath49 ) .
a set of @xmath50 density matrix equations completely describes the dynamics of this space .
however , the ground subspace is coupled to the ground - excited subspace of dimension @xmath51 ( containing states with one excitation as for example @xmath52 ) through the classical field .
this is , in its turn , coupled to the excited subspace of dimension @xmath11 ( containing states of two excitations like @xmath53 ) which can decay back to the ground states . in a density operator approach ,
a total of @xmath54 density matrix elements coupled to each other come into play , which makes the task at hand extremely complex .
some simplifications are possible .
first , terms occurring in the evolution of the ground state density matrix elements are separated into in - terms ( due to spontaneous emission from upper states ) and out - terms ( driving terms due to the presence of the classical field ) , and these terms are treated separately .
second , an amplitude rather than a density matrix approach is sufficient to obtain expressions for the excited - ground and excited subspace density matrix elements which enter the equations .
the procedure is described in appendix b where it is applied to the derivation of the decoherence and population transfer rate for a single 4-level atom .
the treatment is perturbative in the sense that excited state populations are assumed to be negligibly small , as in many treatments of optical pumping .
the states coupled by the fields in this approximation are denoted by @xmath55 , @xmath56 and @xmath57 where the convention used is that the prime indicates excited states ( @xmath58 or @xmath59 ) , while the unprimed symbols indicate ground states ( @xmath60 or @xmath61 ) . by eliminating the intermediate states involving the radiation field ( procedure outlined in refs .
@xcite ) , one can obtain the following coupled equations of motion for the state amplitudes containing one excitation ( @xmath62 and @xmath63 ) , in an interaction picture:@xmath64 in the equation above for @xmath65 , the first term on the right - hand side is the decay ( at a rate @xmath66 equal to half the excited state population decay rate ) of the excited state amplitude of an atom independently coupled to the quantum vacuum .
the second term contains a propagator @xmath67 which includes the effects of the radiation exchange between atoms : an atom makes a transition from state @xmath68 ( quantum number @xmath69 ) to state @xmath70 which is accompanied by a transition in the other atom ( @xmath71 to @xmath72 ) .
the real part of @xmath73 gives a contribution to the decay rate that varies from @xmath44 , for maximum cooperation between atoms when their separation is much less than @xmath30 , to @xmath74 , when no exchange of radiation between atoms is present ( infinite separation ) .
the imaginary part leads to a shift in energy ( which adds to @xmath10 in the equations above ) and varies from @xmath74 ( large separation ) to infinity when the atoms are in the same location .
if the minimum interatomic separation is small but finite , the shift can always be kept small compared to the detuning and can be neglected .
the explicit expressions for the propagators involved in this problem are given in appendix a. the geometrical information on the radiation exchange is contained in the clebsch - gordan coefficients @xmath75 .
the last two terms in the right - hand side arise from driving field induced transitions and from the off - resonant nature of the interaction .
to calculate the out - terms , we note that , as a result of the driving field , ground state amplitudes ( @xmath76 ) are coupled to excited state amplitudes via@xmath77 .
st . amp}%\ ] ] as a result one finds@xmath78 b_{mn~}^{\ast}+\label{out - terms}\\ & + b_{\mu\nu~}\left [ i\chi e^{-i\mathbf{k}_{l}\mathbf{\cdot r}_{1~}% } b_{\alpha n~}\delta_{-1/2,m_{m~}~}+e^{-i\mathbf{k}_{l}\mathbf{\cdot r}_{2~}% } b_{m\alpha~}\delta_{-1/2,m_{n~}~}\right ] .\nonumber\end{aligned}\ ] ] the system of equations ( eqs .
( [ ampeqs ] ) ) is solved for the @xmath51 amplitudes @xmath79 and @xmath80 as functions of the @xmath11 ground state amplitudes @xmath76 ; these expressions are replaced in the above equation and with the identification @xmath81 , the rate equations for @xmath82 are obtained in terms of the @xmath50 density matrix elements @xmath83(with @xmath84 ) .
next , repopulation from the upper states to the lower states is taken into account ( in - terms)@xmath85 where@xmath86 note that some coherence is returned to the ground state as a result of the `` in terms '' .
the derivation of the terms in the right - hand side of eq .
( [ in - terms ] ) is done by tracing over the field states with one - photon occupation number , a procedure which has been used in the case of single multilevel atoms [ see for example ref .
the first two terms describe repopulation and recoherence of the ground manifold from the excited state manifold in a single atom .
cross coupling between atoms is reflected in the next two terms . using again the solutions of eqs .
( [ ampeqs ] ) , the right - hand side of the in - term equations can be expressed in terms of ground state density matrix elements .
a complete system of @xmath50 linear equations is thus obtained by adding the in - term to the out - term contributions .
in this limit , owing to angular momentum conservation rules , the propagators @xmath88 vanish except for @xmath89 .
a few photon exchange processes between atoms are illustrated in fig .
[ exchange1](c ) , along with their accompanying propagators .
notice that due to momentum conservation the polarization of the emitted photon matches the polarization of the absorbed photon .
taking as an example the transfer of excitation from atom @xmath44 in state @xmath90 ( with atom @xmath91 in state @xmath45 ) to atom @xmath91 in state @xmath92 ( with atom @xmath44 in state @xmath93 ) depicted in fig .
[ exchange1 ] , from eq .
( [ in - terms ] ) , one finds that the propagator associated with the exchange is @xmath94 .
[ h ] fig2.eps as in @xcite , the two - atom system can be described by superpositions of states that are either symmetrical or antisymmetrical under particle exchange ( dicke states ) .
the indistinguishability of the particles restricts the system to the symmetric subspace .
the ground state manifold is symmetrized as follows:@xmath95 while ground - excited symmetric states are defined as:@xmath96 rewriting eqs .
( [ ampeqs ] ) in terms of the new coefficients @xmath97and @xmath98 , one finds@xmath99 with quasistatic solutions @xmath100 , \nonumber\\ s_{-1~ } & = 0.\nonumber\end{aligned}\ ] ] notice that the symmetric states containing the excited @xmath59 states have either identically zero or negligibly small amplitude , which substantially simplifies the calculation . in the following two subsections ,
these expressions for the state amplitudes are used to derive the time evolution of the collective coherence and population . in the new basis
the matrix elements are denoted by @xmath101 , with @xmath102 .
the expression for the collective coherence operator expectation value in the new basis is given by@xmath103 in the limit @xmath104,@xmath42the following rate equations are obtained for the out - terms [ from eq .
( [ out - terms])]@xmath105 while the in - terms , obtained from eq .
( [ in - terms ] ) , evolve as@xmath106 adding the out - term contribution to the in - terms , and with the notation @xmath107 ( optical pumping rate ) , the equations for the density matrix elements relevant for the coherence decay are@xmath108 substituting the solutions of the eqs .
( [ coheqs ] ) into eq .
( [ coh ] ) , one obtains@xmath109 this is to be compared with the independent atom coherence decay derived in appendix b [ eq
. b4]@xmath110 [ h ] fig3.eps the increase in the coherence decay rate for intermediate times , as shown in fig .
[ cohdecayx=0 ] , can be understood in terms of the exchange processes in the uncoupled basis illustrated in fig .
[ exchange1 ] . for an independent atom
, the state @xmath90 is reached from the @xmath93 state through the action of the classical field , and decays into @xmath93 and @xmath111with rates @xmath112 and @xmath113 respectively . in the cooperative case ,
other channels responsible for coherence generation or decay appear owing to the presence of the second atom .
an interesting behavior is observed at the initiation stage of the decoherence process , where the coupled system decoheres at a rate equal to that for independent atoms .
however , this is not a general result , but rather a consequence of the initial state prepared with equal ground substate populations .
when both atoms start in the same arbitrary state @xmath114 with @xmath115 and @xmath116 , the evolution of the coupled system coherence takes the following form@xmath117 , \label{arbitrary}%\ ] ] while the independent atom coherence evolves as:@xmath118 expanding the exponentials in eqs .
( [ arbitrary ] ) and ( [ arbitraryindep ] ) for small times @xmath119 , we find@xmath120 \text{,}%\end{aligned}\ ] ] which shows that the decoherence rate of the coupled system is modified by the term @xmath121 .
this indicates that , given a population imbalance between the up and down states , at the initiation stage , the decoherence rate of the coupled system can be either larger or smaller than the one for the independent atoms , and vanishes for the balanced case only , when @xmath122 .
[ h ] fig4.eps in the symmetric basis , the expectation value of the collective population operator is expressed as@xmath123 three density matrix elements are coupled to each other : @xmath124 , @xmath125and @xmath126 .
the evolution resulting from the classical field can be obtained [ from eq .
( [ out - terms ] ) ] as@xmath127 while the in - terms are given by [ see eq .
( [ in - terms])]@xmath128 combining the in - terms with out - terms , one finds rate equations@xmath129 that are solved to give@xmath130 this is to be compared with the independent atom population evolution@xmath131 where only one mechanism for populating state @xmath132 is present : excitation of state @xmath92 by the classical field followed by decay at a rate @xmath133 to state @xmath134 .
[ h ] fig5.eps the increase in the population transfer rate , plotted in fig .
[ opticalpumpingx=0 ] vs. the independent atoms rate , is due to photons emitted by the second atom that drive the first atom into the @xmath18 state .
a faster excitation of state @xmath135leads to a faster transfer to @xmath132 for intermediate times .
it is interesting to isolate the dynamics of a single atom with the purpose of identifying the mechanisms that lead to the modification of its radiative properties due to the presence of a second atom in its vicinity .
we analyze the rate of change of the expectation value of the one atom coherence operator @xmath136 .
this is expressed in terms of density matrix elements as @xmath137 , and it is found to satisfy the following equation of motion [ using eqs .
( [ out - terms ] ) and ( [ in - terms])]@xmath138 .
\label{oneatcoh}%\ ] ] the first term in the right - hand side in the above equation simply indicates the decay of the coherence of the independent atom at the expected rate @xmath139 .
the second term contains the modification induced by the action of the neighboring atom . at the moment when the interaction between atoms is initiated
, the density matrix of the coupled system can be factorized and this term can be written as : @xmath140 , where @xmath141 is the population difference operator for the first atom .
the significance of this term is that @xmath2 polarization ( coherence ) established in the second atom induces @xmath2 polarization in the first atom through the vacuum , given a population difference .
we proceed now to analyze the origin of this coupling term by considering two distinct situations in which atom @xmath91 is prepared in a superposition @xmath142 exhibiting coherence equal to @xmath143 , while atom @xmath44 is prepared either in the @xmath134 state or the @xmath12 state . in both cases
no initial @xmath2 atomic polarization in the atom of interest is present . when starting with the population in the @xmath134 state , using eqs .
( [ out - terms ] ) and ( [ in - terms ] ) , expressions for the density matrix elements present in the right - hand side of eq .
( [ oneatcoh ] ) can be derived , and an out - term is found to be responsible with the generation of coherence@xmath144 the process leading to this can be represented as follows@xmath145 where in the first step , the field induces an excitation in the second atom , followed by a swap of excitation between atoms through the vacuum ( collision - like effect ) and a stimulated emission from the first atom . in the second case , where the population is initially stored in the @xmath45 state , both an
in - term and an out - term are present : @xmath146 the coherence generation through the out - term is similar to the process shown above [ eq .
( [ upprocess ] ) ] , while the in - term takes the following path@xmath147 where consecutive excitations for both atoms are followed by spontaneous decay into a state with coherence in the first atom .
simple analytical results are not available in this regime . coupling through propagators other than @xmath148 takes place .
the polarization of the emitted photon does nt have to match the one of the absorbed photon , a situation which is illustrated in fig .
[ exchange2 ] , where the coupling of the two - atom state @xmath149 to other states is shown and the corresponding elements of the @xmath73 matrix ( including non - diagonal ones ) responsible for the coupling specified .
[ h ] fig6.eps the calculations are now performed in the uncoupled basis .
numerical solutions of @xmath150 coupled rate equations give the coherence decay , whereas @xmath151 rate equations are solved to obtain the population transfer rate .
the output of our numerical simulations is dependent both on time and on the spherical coordinates of the second atom @xmath152and @xmath153 .
since the system has azimuthal symmetry , the interesting cases are obtained by varying @xmath154 and @xmath155 .
two orientations of the system with respect to the field propagation are examined : @xmath156@xmath157(@xmath158 ) and @xmath159@xmath157(@xmath160 ) .
the @xmath11 density matrix elements responsible for the coherence [ eq .
( [ cohandpop ] ) ] , are coupled to @xmath51 more through nondiagonal elements of @xmath73 .
a system of @xmath150 linear differential equations has to be solved containing @xmath161 .
the expressions for all @xmath150 rate equations are not given here ; instead a single one is written to illustrate the way the coupling among different states comes into play@xmath162{c}% -5\rho_{\uparrow\downarrow;\downarrow\downarrow~}-e^{-ik_{l}r_{21}\cos\theta } \left ( \begin{array } [ c]{c}% \sqrt{2}g_{10~}(r_{21},\theta,\varphi)\rho_{\downarrow\downarrow ; \downarrow\downarrow~}+2g_{00~}(r_{21},\theta,\varphi)\rho_{\downarrow \uparrow;\downarrow\downarrow~}\\ + 2g_{11~}(r_{21},\theta,\varphi)\rho_{\uparrow\downarrow;\downarrow \downarrow~}% \end{array } \right ) \\ + e^{ik_{l}r_{21}\cos\theta}\left ( \begin{array } [ c]{c}% 2g_{11~}(r_{21},\theta,\varphi)\rho_{\uparrow\downarrow;\downarrow\downarrow ~}+\sqrt{2}g_{01~}(r_{21},\theta,\varphi)\rho_{\uparrow\downarrow ; \downarrow\uparrow~}\\ + \sqrt{2}g_{01~}(r_{21},\theta,\varphi)\rho_{\uparrow\downarrow;\uparrow
\downarrow~}% \end{array } \right ) \end{array } \right ] .\ ] ] [ h ] fig7.eps the time evolution of the collective coherence is shown in fig .
[ intermediatecohdecay ] for @xmath163 . for both @xmath164@xmath165 and @xmath166@xmath165
, the coherence decays more rapidly than for the independent atom case , and the decay for @xmath156@xmath165 is faster than that for @xmath166@xmath165 . to obtain some idea of the dependence of the decay rate on interatomic separation , we plot in fig .
[ ratevardistance ] the decoherence rate ( @xmath167 ) as a function of distance , at a fixed time @xmath168 , for both the @xmath169and @xmath170 . even though a numerical solution has been used to obtain the plot in fig .
[ ratevardistance ] , an approximate analytical treatment can provide insight into the qualitative nature of the results . in particular it can help explain why the parallel case decay rate is larger than that for closely separated atoms .
the coupled equations of motion for the coherence operators associated with each atom [ similar to eq .
( [ oneatcoh ] ) but with the difference that now the coupling coefficients are @xmath171dependent ] are solved approximately at a fixed time . it is found that the perpendicular case differs from the close atoms case only insofar as the coupling between atoms is modulated by the real part of @xmath172 . an analytical expression for a fixed time @xmath173 gives a decoherence rate of the collective coherence that varies with the separation as @xmath174 $ ] .
there is , however , a fundamental difference between the perpendicular and parallel case that is reflected in the system s response . in the perpendicular case
the spatial phase of the laser field does not enter since @xmath175 . as a result
the spin coherence associated with each atom evolves in an identical fashion . on the other hand , in the parallel case
there is a relative phase difference of @xmath176 associated with the laser field as it interacts with atoms at the two sites .
consequently , the response of the two atoms need no longer be identical , since the spatial symmetry has been broken by the field .
three coupling terms are present in the parallel case : @xmath177 , @xmath178and @xmath179 ; the importance of the laser induced spatial phase is evident in these expressions . owing to the extra couplings , the two atoms accumulate different spatial phases , and the collective coherence ( obtained as the sum of individual coherences ) shows a spatial modulation that varies as @xmath180 .
a full analytical solution for the variation of the decay rate with the distance for any fixed time is not available ; however , a perturbative treatment for small times ( @xmath119 ) indicates an increase in the decay rate @xmath181 @xmath182 .
[ h ] fig8.eps just as in the case of close atoms , we extend our simulations to analyze the decay of an arbitrary initial state @xmath183 .
the observed behavior is quite different here .
even for relatively large separation ( @xmath184 ) , a state with most population in the down state decays much faster than in the independent atom case , while an inhibition of decoherence is obtained when the initial state is prepared with more population in the up state [ as shown in fig .
[ abstatex=1and2.5 ] for @xmath156@xmath165 and @xmath166@xmath165 ] .
[ h ] fig9.eps one more density matrix element ( @xmath186 ) is coupled to the @xmath150 listed before and a system of @xmath151 rate equations is solved to obtain the collective population transfer as a function of time .
[ h ] fig10.eps an interesting behavior is obtained in both the parallel and perpendicular cases ( see fig .
[ intermediatepoptransfer ] ) : the optical pumping rate is initially larger and afterwards smaller than the one for independent atoms .
the saturation effect at large times is due to the addition of a decay channel ( the @xmath187state ) which provides a way for the transfer of population from @xmath134 back to the initial @xmath93 state .
this accounts for a slow down at large times where the repopulation from the second atom ( resulting in population in state @xmath188 ) is comparable with the population produced by the classical field .
as seen in sec .
iii b , a partial transfer of coherence from an @xmath189polarized atom to an initially unpolarized atom ( in a @xmath3 state , either up or down ) can be achieved .
the two distinct situations discussed there can be extended for variable interatomic separations .
figures [ down ] and [ up ] show the evolution of the coherence of the first atom as a function of time for @xmath190 and @xmath184 and @xmath191 , respectively , when @xmath156@xmath165 . in both cases , owing to the oscillating nature of the coupling between atoms ( as a function of separation ) the sign of the effect produced in the initially @xmath3 polarized atom
depends dramatically on the distance .
[ h ] fig11.eps the decay dynamics are still given by eqs .
( [ upeq ] ) and ( [ downeq ] ) , with the distinction that the decay parameter , which for close atoms is simply equal to @xmath192 , is now spatially modulated by @xmath193 . in the case where the first atom is initially in the up state , the coherence is driven only by the out - term , which leads to a simple time behavior , where @xmath194 changes sign only due to the spatial oscillation of the coupling term .
in contrast , in the down case , the competition between the in - term and out - term leads to a change in the sign of @xmath195 for intermediate distances .
[ h ] fig12.eps
we examined a system of two multilevel atoms coherently driven by a single - mode classical laser field and coupled to the electromagnetic vacuum .
the decoherence ( of a quantum superposition stored in the ground sublevels ) that necessarily accompanies the process of manipulation of the atomic states has been analyzed in the context of cooperative behavior .
two cases have been treated , where the atoms are either at the same position or separated by a distance comparable to the optical wavelength .
it has been found ( not surprisingly ) that , for the case of close atoms , the `` communication '' between atoms leads to an increase in both the decoherence and population transfer rates . with increasing interatomic separation , in the case of the field propagating perpendicularly to the line joining the atoms ,
the decoherence rate is less than that for close atoms .
this is an expected result since , from the point of view of the classical field the atoms are located at equivalent positions , and a simple decrease of the interatomic coupling due to the increasing separation is expected . for certain distances , however , owing to the oscillating behavior of this coupling , a small effect of decoherence inhibition is also observed .
a more interesting situation arises for the case when the atoms are aligned parallel to the field propagation direction . the equivalency of positions does not hold here anymore , and a spatial phase difference between atoms resulting from the classical field is established .
the coupling through the vacuum is modulated by this spatial phase difference , and a considerable enhancement in the decay rate is observed at separations of order @xmath196 .
these results will be generalized in a future planned publication to a large ensemble of atoms in a pencil - shaped geometry .
in such a medium , at fresnel numbers close to unity , the atoms are practically aligned along the direction of the field ; for large optical depths , the phase effect described above is expected to lead to a substantial increase in the decay of the collective atomic coherence .
this work is supported by the national science foundation under grant no .
phy-0244841 and the focus center grant .
the expressions for the propagators involve spherical harmonics and hankel functions of the first kind , @xmath197 and @xmath198 [ ref .
we explicitly derive the equations of motion for the ground state coherence and populations for a single 4-level atom interacting with a @xmath1 polarized field . in the perturbative limit where a maximum of one excitation is allowed ,
a basis set in the hilbert space of atom and radiation field is comprised of states @xmath200 , @xmath201 , @xmath202 , @xmath203 , @xmath204 and @xmath205 . the first ket denotes the state of the atom while , the second one describes a vacuum field or a one photon state with wave number @xmath7 and polarization @xmath30 .
the state amplitudes obey the following equations of motion : @xmath206 using a master equation approach one can write density matrix equations of motion for the ground state sublevels as:@xmath207 the first observation that we make here is that the terms in the right hand side of the above equations , that are due to the field ( out - terms ) and to the coupling to the vacuum ( in - terms ) can be derived separately .
in other words , the presence of dissipation can be neglected when writing equations describing the driving effect of the field and later added to the equations phenomenologically .
the derivation of rate equations can now be carried out by writing equations for the ground - excited coherences and excited state populations and coherences and adiabatically eliminating them . however
, this is an unnecessary complication ; the second observation that we make provides an easier way of doing this . only the first @xmath11 equations in eqs .
( b1 ) have to be solved and the excited amplitudes can be written in terms of ground state amplitudes .
next , the density matrix elements can be written simply as if the state of the system were a pure state ( as products of amplitudes ) and replaced in eqs .
( 2 ) to obtain rate equations .
the separation of in - terms from out - terms in these equations insures the validity of this approach .
following this recipe , it is found that ( in the limit @xmath208)@xmath209 replacing these expressions into eqs .
( b2 ) , rate equations are obtained in a final form:@xmath210 the decoherence of an initial state @xmath211 can now be calculated as described by the evolution of @xmath212(neglecting the phase associated with the ac stark shift ) @xmath213 while the population transfer from state @xmath93 to state @xmath134 is given by@xmath214 l. k. thomsen , s. mancini and h. m. wiseman , phys .
rev a * 65 * , 061801 ( 2002 ) ; s. massar and e. s. polzik , phys . rev . lett . *
91 * , 060401 ( 2003 ) ; j. k. stockton , r. van handel and h. mabuchi , phys .
a * 70 * , 022106 ( 2004 ) ; c. genes and p. r. berman , phys . rev .
a * 73 * , 013801 ( 2006 ) .
it should be noted that the qualitative results for this particular case can not be generalized to other level scheme or field polarization .
for example , a calculation on two close atoms ( not presented here ) that uses @xmath2 instead of @xmath1 polarized driving fields and the same @xmath0 transition , indicates that the collective decoherence rate is not different from the independent atom coherence decay rate . | an ensemble of multilevel atoms is a good candidate for a quantum information storage device .
the information is encrypted in the collective ground state atomic coherence , which , in the absence of external excitation , is decoupled from the vacuum and therefore decoherence free .
however , in the process of manipulation of atoms with light pulses ( writing , reading ) , one inadvertently introduces a coupling to the environment , i.e. a source of decoherence .
the dissipation process is often treated as an independent process for each atom in the ensemble , an approach which fails at large atomic optical depths where cooperative effects must be taken into account . in this paper , the cooperative behavior of spin decoherence and population transfer for a system of two , driven multilevel - atoms is studied . not surprisingly , an enhancement in the decoherence rate is found , when the atoms are separated by a distance that is small compared to an optical wavelength ; however , it is found that this rate increases even further for somewhat larger separations for atoms aligned along the direction of the driving field s propagation vector .
a treatment of the cooperative modification of optical pumping rates and an effect of polarization swapping between atoms is also discussed , lending additional insight into the origin of the collective decay . |
let @xmath2 be a _ pure noetherian complete algebra _ ,
i.e. an associative @xmath3-algebra such that : 1 .
its center @xmath4 is a complete local noetherian @xmath3-algebra .
@xmath2 is finitely generated @xmath4-module without minimal submodules .
denote by @xmath5 the radical of @xmath2 .
it was shown in @xcite that @xmath2 is tame if and only if it satisfies the following conditions : 1 .
the algebra @xmath6 is hereditary .
2 . @xmath7 .
3 . for any simple left @xmath2module @xmath8
the length of the left @xmath2-module @xmath9 is at most 2 and for any simple right @xmath2module @xmath10 the length of the right @xmath2module @xmath11 is at most 2 .
we call algebras satisfying these conditions _ nodal _ algebras .
our description of the derived category of a nodal algebra shows that it is tame at least in `` pragmatic '' sense , i.e. one can obtain a list of its indecomposable objects as a union of one - dimensional families and some discrete set of objects staying apart .
unfortunally , the definition of derived tameness proposed in @xcite can be only applied to finite dimensional algebras of finite global dimension and nodal algebras usually satisfy neither of these conditions .
the methods developed in this article can be also applied to finite dimensional gentle and skew - gentle algebras considered in @xcite , @xcite , @xcite , @xcite , @xcite , as well as to some other algebras @xcite and to some derived categories of coherent sheaves @xcite .
an advantage of these methods is that they also work in cases , when an algebra has infinite homological dimension and describe the derived category of bounded from the _ right _ complexes .
the developed technique allows to write down projective resolutions of indecomposable complexes .
for the sake of simplicity we suppose that the field @xmath3 is algebraically closed .
let us rewrite the definition of nodal algebras in a more transparent form .
let @xmath8 be a simple @xmath2-module , @xmath12 its projective covering .
then we have an exact sequence @xmath13 apply the functor @xmath14 to this sequence .
we get @xmath15 but @xmath16 , hence @xmath17 .
so we have an exact sequence @xmath18 therefore @xmath19 is a direct sum of simple @xmath20-modules . let @xmath21 be the set of all non - isomorphic simple @xmath2-modules , @xmath22 the set of all non - isomorphic simple @xmath20-modules . consider the graph @xmath23 with vertices @xmath24 , @xmath25 .
there is an arrow from @xmath26 to @xmath27 if and only if @xmath27 is a direct summand of @xmath28 .
then , as it was shown in @xcite , the last condition in the criteria of tameness is equivalent to the following condition : all connected components of @xmath23 are of the form : 1 .
@xmath29 2 .
@xmath30 3 .
@xmath31 let us consider some examples .
let @xmath32/(xy)$ ] , @xmath33 be its maximal ideal
. then @xmath34\times { \mathbf k}[[y]]$ ] .
let @xmath8 be the unique simple @xmath2-module , @xmath35 be simple @xmath20-modules .
then the graph @xmath23 has the form @xmath36 let @xmath37 , 1\le i , j \le 2 ; f_{11}(0 ) = f_{22}(0 ) \ }
\subseteq { \rm mat}_{2}({\mathbf k}[[t]]).\ ] ] as one can easily observe , @xmath2 is just the algebra @xmath38 ( @xmath39 is the algebra of formal power series in two non - commutative variables ) .
the endomorphism algebra of its radical is just @xmath40 , 1\le i , j \le 2 \ } \subseteq { \rm mat}_{2}({\mathbf k}[[t]]).\ ] ] it is easy to see that it is just the completed path algebra of the quiver indeed , an isomorphism is given by @xmath41 the graph @xmath23 again has the form @xmath36 here and further on we consider the natural completion of path algebras , namely , the @xmath42-adic one , where @xmath42 is the ideal generated by all arrows .
let @xmath2 be the completed path algebra of the gelfand quiver as one can easily see , @xmath43 , 1\le i , j \le 3 \ } \subseteq { \rm mat}_{3}({\mathbf k}[[t]])\ ] ] an isomorphism is given by @xmath44 @xmath45 the endomorphism algebra @xmath46 is @xmath47 , 1\le i , j \le 3 \ }
\subseteq { \rm mat}_{3}({\mathbf k}[[t]]).\ ] ] and is morita equivalent to @xmath48 , 1\le i , j \le 2 \ } \subseteq { \rm mat}_{2}({\mathbf k}[[t]]),\ ] ] which is the completed path algebra of the quiver note that @xmath20 is isomorphic to the completed path algebra of the following ( non - basic ) quiver let @xmath49 be simple @xmath2-modules , @xmath50 simple @xmath20-modules .
our graph @xmath23 has the form @xmath51
let @xmath2 be a semi - perfect associative @xmath3-algebra ( not necessarily finite dimensional ) , @xmath52 be an embedding such that @xmath53 .
let @xmath54 be a two - sided @xmath20-ideal containing @xmath5 .
it means that @xmath55 , thus @xmath56 and @xmath57 are semi - simple algebras .
let @xmath14 be the derived functor of the tensor product .
we want to describe the fibers of the map @xmath58 @xmath59 denotes the category of _ finitely - generated _ @xmath2-modules .
we always consider objects of derived categories as complexes of projective modules .
consider the following category of triples of complexes @xmath60 1 .
objects are triples @xmath61 , where + @xmath62 , + @xmath63 , + @xmath64 a morphism in @xmath65 , such that + @xmath66 is an isomorphism in @xmath67 2
. morphisms @xmath68 are pairs @xmath69 , @xmath70 such that + [ cols="<,^ " , ] this complex is isomorphic to a projective resolution of a module which is finitely generated but not finite dimensional .
there are finally complexes which are not coming from the matrix problem .
they are just complexes of the form @xmath71 , which come from triples + @xmath72 .
the description of complexes for a general nodal algebra can be obtained by combining the combinatorics of complexes of the derived category of the dihedral algebra and of the gelfand quiver .
in @xcite it was proven that there are only two cases of compact lie groups , for which the category of harish - chandra modules is tame : @xmath73 and @xmath74 . as a corollary of the theorem
we obtain that the derived category of harish - chandra modules is also tame in both of these cases .
we have already seen it for @xmath73 .
let @xmath75 be the connected component of @xmath74 . 1 .
let @xmath76 .
then the category of harish - chandra modules over @xmath75 at a singular point is equivalent to the category of finite - dimensional representations of the completed path algebra of the following quiver ( see @xcite ) : + + where we have the relations : @xmath77 and @xmath78 is nilpotent , @xmath79 @xmath80 moreover , all @xmath81 are nilpotent . + this algebra can be embedded into + + ( this algebra is the endomorphism algebra of the radical of @xmath2 .
however , it is not so important ) .
+ the simple @xmath2-module , corresponding to the vertex @xmath82 is of the first type . those , which correspond to @xmath83 are of the second type , @xmath84 are of the third type .
2 . if @xmath85 .
then the category of harish - chandra modules over @xmath75 is described by the completed path algebra of the following quiver : + + with relations : @xmath86 @xmath87 @xmath88 is nilpotent , @xmath89 and all @xmath90 are nilpotent .
+ it can be embedded into + + the simple module , corresponding to the vertex @xmath91 is of the first type , all other simple modules are of the second type .
let us consider two more examples ( see @xcite for a description of indecomposable modules over these algebras ) .
consider the completed path algebra of the following quiver : where we have the relations @xmath92
@xmath93 @xmath94 and finally @xmath95 is nilpotent . as one easily observes , we can embed this algebra into the simple @xmath2-modules , corresponding to vertices @xmath96 are of the second type , @xmath97 are of the third type .
consider the completed path algebra of quiver : where we have the relations : @xmath98 and @xmath99 are nilpotent , @xmath100 @xmath101 @xmath102 moreover , all @xmath103 are nilpotent .
we can embed this algebra into : the simple @xmath2-modules , corresponding to vertices @xmath96 are of the second type , @xmath97 are of the third type .
it can be checked that all algebras from this section are nodal and every embedding is embedding into the endomorphism algebra of the radical . * appendix *
# 1 # 1 # 1#2#3#1#2 [ sec1a ] let @xmath104 be an additive category .
we denote by @xmath105 the category of chain complexes with entries from @xmath104 and by @xmath106 the factorcategory of @xmath105 modulo homotopy .
if @xmath104 is abelian , we denote by @xmath107 the _ derived category _ of @xmath104 , that is the category of quotients of @xmath106 with respect to the set of morphisms inducing isomorphism of homologies .
we fix a commutative ring @xmath108 and consider @xmath109-_categories _
, namely , suppose that all sets @xmath110 are modules over @xmath108 and the multiplication of morphisms is @xmath111-bilinear .
[ 1a1 ] an additive category @xmath104 is called called : * _ local _ if every object @xmath112 decomposes into a finite direct sum of objects with local endomorphism rings ; * _ @xmath113-local _ if every object @xmath112 decomposes into a finite of countable direct sum of objects with local endomorphism rings ; * _ fully additive _ if any idempotent morphism in @xmath104 splits , that is defines a decomposition into a direct sum ; * _ _ ( over @xmath109 ) if it is fully additive and all morphism spaces @xmath114 are finitely generated @xmath109-modules .
especially if @xmath109 is a field , a category is called _ locally finite dimensional_. evidently , every category is local ; moreover , an endomorphism algebra @xmath115 in a category is a _ finite @xmath109-algebra _
, i.e. such that the underlying @xmath109-module is finitely generated .
it is known that any local ( or @xmath113-local ) category is fully additive ; moreover , a decomposition into a direct sum of objects with local endomorphism rings is always unique ; in other words , any local ( or @xmath113-local ) category is a _ krull schmidt _ one ,
* theorem 3.6 ) . for
a local category @xmath104 denote by @xmath116 its _ radical _ , that is the set of all morphisms @xmath117 , where @xmath118 , such that no component of the matrix presentation of @xmath119 with respect to some ( hence any ) decomposition of @xmath120 and @xmath121 into a direct sum of indecomposable objects is invertible .
note that if @xmath122 , there is a morphism @xmath123 such that @xmath124 and @xmath125 .
hence both @xmath126 and @xmath127 are nonzero idempotents , which define decompositions @xmath128 and @xmath129 such that the matrix presentation of @xmath119 with respect to these decompositions is diagonal : @xmath130 , and @xmath131 is invertible .
obviously , if @xmath104 is locally finite dimensional , then @xmath132 coincide with the set of all morphisms @xmath117 such that @xmath126 ( or @xmath127 ) is nilpotent for any morphism @xmath123 . [ 1a2 ] suppose that @xmath109 is a complete local noetherian ring with maximal ideal @xmath133 . if @xmath104 is a category over @xmath109 , the categories @xmath105 and @xmath106 are @xmath113-local ( in particular , krull schmidt ) .
moreover , a morphism @xmath134 from one of these categories belongs to the radical all components @xmath135 ( or @xmath136 ) are nilpotent modulo @xmath133 for any morphism @xmath137 . proof .
denote by @xmath138 the residue field of @xmath109 .
we use the following simple statement from linear algebra .
[ 12 ] let @xmath139 be a finite dimensional @xmath140-algebra and @xmath141 be and element from @xmath139 .
there is a polynomial @xmath142\,$ ] such that @xmath143 is an idempotent and @xmath144 for any idempotent @xmath145 from any @xmath146-algebra .
moreover , @xmath143 is nilpotent ( or invertible ) so is @xmath141 . proof .
suppose that a polynomial @xmath147 satisfies the condition @xmath148 , @xmath149 .
then @xmath150 for any idempotent @xmath151 from any finite dimensional algebra .
we can embed @xmath139 in an endomorphism algebra of some finite dimensional vector space @xmath152 , so we suppose that @xmath153
. decompose @xmath154 so that the restriction @xmath155 is nilpotent and @xmath156 is invertible . replacing @xmath141 by @xmath157 for some @xmath158 , one can suppose that @xmath159 .
indeed , if we have found a polynomial @xmath147 such that @xmath160 is idempotent , then @xmath161 hence @xmath162 is the polynomial for @xmath163 we are looking for . in particular , if @xmath164 , then we can take @xmath165 . set @xmath166 since @xmath167 is invertible , there is a polynomial @xmath168 such that @xmath169 and @xmath170 . if @xmath171 is an eigenvalue of @xmath167 , then @xmath172 , whence @xmath173 for every idempotent @xmath145 .
if @xmath171 is not an eigenvalue of @xmath167 , then @xmath174 , where @xmath175 is the minimal polynomial of @xmath167 , hence there is a polynomial @xmath176 such that @xmath177 and @xmath178 .
therefore , @xmath179 and @xmath144 for every idempotent @xmath145 , which accomplishes the proof of the lemma .
@xmath180 recall also a known result , which can be easily deduced , for instance , from ( * ? ? ? * section iii.8 ) .
there are polynomials @xmath181 $ ] with @xmath182 and such that for every ring @xmath183 , any ideal @xmath184 and any element @xmath185 such that @xmath186 , @xmath187 and @xmath188 .
( for instance , @xmath189 . )
[ ca1 ] let @xmath183 be a finite algebra over a local noetherian ring @xmath109 with maximal ideal @xmath133 and @xmath185 .
for every @xmath190 there is a polynomial @xmath191 $ ] such that * @xmath192 ; * @xmath193 for every element @xmath151 of an arbitrary finite @xmath109-algebra such that @xmath194 ; * @xmath195 @xmath88 is invertible ; * @xmath196 @xmath88 is nilpotent modulo @xmath133 . proof .
find a polynomial @xmath147 over @xmath138 such that @xmath197 is idempotent in @xmath198 , where @xmath199 and @xmath200 for any idempotent @xmath201 of any @xmath146-algebra ; especially @xmath148 .
lift @xmath147 to a polynomial @xmath202 $ ] such that @xmath203
. then @xmath204 is idempotent modulo @xmath133 and if @xmath194 , then @xmath205 ( by the construction of @xmath206 ) and @xmath207 ( it is true for any polynomial @xmath206 satisfying @xmath208 ) .
set @xmath209 .
then @xmath210 is idempotent modulo @xmath211 , just as @xmath212 for every @xmath151 that is idempotent modulo @xmath133 . if , moreover , @xmath194 , then @xmath213 and @xmath214 .
let @xmath215 ; then @xmath216 and @xmath217 .
therefore it holds @xmath218 , wherefrom @xmath219 .
but then @xmath220 , so @xmath221 .
@xmath180 let now @xmath222 be an endomorphism of a complex @xmath223 from @xmath105 .
consider the sets @xmath224 defined as follows : @xmath225 , @xmath226 and @xmath227 .
obviously , @xmath228 , @xmath229 and @xmath230 consists of a unique element @xmath231 . using corollary [ ca1 ]
, we can construct a sequence of endomorphisms @xmath232 such that * @xmath233 ; * @xmath234 ; * @xmath235 is invertible or nilpotent modulo @xmath133 so is @xmath236 .
then one easily sees that setting @xmath237 , we get an idempotent endomorphism @xmath238 of @xmath239 , such that @xmath240 @xmath236 is nilpotent modulo @xmath133 ( respectively @xmath236 is invertible ) .
especially , if either one of @xmath241 is neither nilpotent nor invertible modulo @xmath133 or one of @xmath241 is nilpotent modulo @xmath133 while another one is invertible , then @xmath242 is neither zero nor identity .
hence the complex @xmath223 decomposes .
thus @xmath223 is indecomposable if and only if , for any endomorphism @xmath222 of @xmath223 , either @xmath222 is invertible or all components @xmath243 are nilpotent modulo @xmath133 . since all algebras @xmath244 are finite dimensional , neither product @xmath245 where @xmath246 and one of them is nilpotent modulo @xmath133 , can be invertible
therefore , the set of endomorphisms @xmath222 of an indecomposable complex @xmath223 such that all components @xmath243 are nilpotent modulo @xmath133 form an ideal @xmath247 of @xmath248 and @xmath249 is a skew field .
hence @xmath250 and @xmath248 is local .
now we want to show that any complex from @xmath105 has an indecomposable direct summand . consider an arbitrary complex @xmath223 and suppose that @xmath251 . for any idempotent endomorphism @xmath252 of @xmath239
at least one of the complexes @xmath253 or @xmath254 has a non - zero component at the zero place . on the set of all endomorphisms of @xmath223 we can introduce a partial ordering by writing @xmath255 @xmath256 and both @xmath257 and @xmath258 are non - zero .
let @xmath259 be a chain of idempotent endomorphisms of @xmath239 . as all endomorphism algebras @xmath260 are finitely generated @xmath109-modules , the sequences @xmath261 stabilize for all @xmath91
, so this chain has a lower bound ( formed by the limit values of components ) . by zorn s lemma , there is a minimal non - zero idempotent of @xmath239 , which defines an indecomposable direct summand .
again , since all @xmath262 are finitely generated , for every @xmath263 there is a decomposition @xmath264 where all @xmath265 are indecomposable and @xmath266 for @xmath267 .
moreover , one may suppose that @xmath268 for @xmath269 and @xmath270 for @xmath271 .
evidently , it implies that @xmath272 where @xmath273 and @xmath274 for @xmath271 , which accomplishes the proof of the proposition [ 1a2 ] for @xmath105 .
note now that the endomorphism ring of each complex @xmath275 in the category @xmath106 is a factor - ring of its endomorphism ring in @xmath105 .
hence it is either local or zero ; in the latter case the image of @xmath275 in @xmath106 is a zero object .
therefore , the claim is also valid for @xmath106 .
@xmath180 [ 1a3 ] let @xmath109 be local , complete and noetherian , and @xmath276 be an @xmath109-algebra finitely generated as @xmath109-module .
then the derived category @xmath277 , where @xmath278 denotes the category of finitely generated @xmath276-modules , is @xmath279-local , in particular , krull
schmidt category .
proof . indeed , @xmath277
coincide with the category @xmath280 , where @xmath281 denotes the category of finitely generated projective @xmath276-modules .
@xmath180 _ the conditions of proposition [ 1a2 ] are essential indeed , and krull
schmidt theorem can fail even for the category of bounded complexes @xmath282 over a local category @xmath283 as the following example shows .
_ let @xmath284 be the localization of the polynomial ring @xmath285\,$ ] at the maximal ideal @xmath286 , @xmath287 be the category of free @xmath288-modules of finite rank . obviously , it is local .
the category @xmath289 is equivalent to the category @xmath290 and contains the category @xmath291 as a full subcategory .
denote by @xmath108 the factor - ring @xmath292 .
it is a local domain , but its completion @xmath293 is not a domain : its normalization decomposes as @xmath294 , where each @xmath295\,$ ] . in particular
, @xmath293 has three torsion - free modules @xmath296 such that each @xmath296 has a composition series with the factors @xmath297 , where @xmath298 ( it is the projection of @xmath299 . onto @xmath300 ) .
it implies that @xmath108 has torsion - free indecomposable modules @xmath301 with the following completions : @xmath302 ( cf .
then @xmath303 , hence the category @xmath304 , thus also @xmath305 and @xmath306mod ) are not local .
we summarize the results of @xcite , changing both the definition and the presentation of the answer to equivalent ones , which seem more convenient for our purpose .
moreover we add a description of some infinite dimensional representations that occur in dealing with derived categories , together with a sketch of a proof .
as v.m.bondarenko has informed us , he has submitted a paper containing more details on infinite case @xcite .
note that bondarenko calls `` bunch of semichained sets '' what we call `` bunch of chains '' .
the reason can be seen if one compares our definitions .
[ a11 ] a _ bunch of chains _ @xmath307 consists of : * an index set @xmath308 , which we suppose finite or countable . * for each @xmath309 , two chains ( linearly ordered sets ) @xmath310 and @xmath311 .
+ we set @xmath312 and @xmath313 . * a symmetric relation @xmath314 ( not an equivalence ! ) on @xmath315 such that for every @xmath316 there is at most one @xmath317 with @xmath318 ( maybe @xmath319 ) .
we define an equivalence relation @xmath320 on @xmath315 such that @xmath321 means either @xmath319 or @xmath318 , and set @xmath322 .
we write @xmath323 if there is an index @xmath309 such that @xmath324 or vice versa . for each @xmath325 such that @xmath326 we introduce two new elements @xmath327 and set @xmath328 .
we subdivide @xmath329 into @xmath330 and @xmath331 , which are the images of @xmath310 and @xmath311 ; for instance @xmath332 and @xmath333 are in @xmath334 if @xmath335 .
we consider the ordering @xmath336 on @xmath315 , which is just the union of orderings on all @xmath310 and @xmath311 , and extend it , as well as the relation @xmath337 , onto @xmath329 so that each `` new '' element @xmath332 or @xmath333 inherits all relations that the element @xmath316 has .
for instance , @xmath338 with @xmath339 means that @xmath340 ; @xmath341 means that @xmath342 , etc .
note that the elements @xmath327 are always non - comparable . on the other hand , we extend the equivalence @xmath320 to @xmath329 trivially ( each new element @xmath332 or @xmath333 is unique in its @xmath320-class ) , and set @xmath343 . a bunch of chains @xmath307 gives rise to a bimodule problem .
namely , we fix a field @xmath146 and define a @xmath146-category @xmath344 and an @xmath276-bimodule @xmath345 as follows : * @xmath346 . * if @xmath347 are two equivalence classes , a basis of the morphism space @xmath348 consists of elements @xmath349 with @xmath350 and , if @xmath351 , the identity morphism @xmath352 .
* the multiplication is given by the rule : @xmath353 if @xmath354 , while all other possible products are zeros .
* a basis of @xmath355 consists of elements @xmath356 with @xmath357 . *
the action of @xmath276 on @xmath358 is given by the rule : @xmath359 if @xmath360 ; @xmath361 if @xmath362 , while all other possible products are zeros .
the category of _ representations of the bunch _ @xmath363 over the field @xmath146 is then defined as the category @xmath364 of the elements of this bimodule . in other words ,
a representation is a set @xmath365 of block matrices @xmath366 such that @xmath321 implies @xmath367 .
two representations are isomorphic they can be obtained from one another by a sequence of the following _ elementary transformations _ : * elementary transformations of rows ( columns ) in each horizontal ( vertical ) stripe ; it means that they are performed simultaneously in all matrices @xmath368 with fixed @xmath316 ( respectively @xmath317 ) ; moreover , if @xmath369 , the transformations of the @xmath316-stripe must be the same as those of @xmath370-stripe ( certainly , if one of them is horizontal and the other is vertical , `` the same '' means `` contragredient '' ) ; * if @xmath340 , then scalar multiples of rows ( columns ) of the @xmath316-stripe can be added to rows ( columns ) of the @xmath317-stripe .
one easily sees that this definition coincides with that of @xcite .
the description of indecomposable representations from @xcite rests upon a combinatorics , which we expound in terms of _ strings and bands _ alike to their use in the representation theory .
[ a12 ] let @xmath371 be a bunch of chains
an @xmath363-_word _ is a sequence @xmath372 , where @xmath373 and @xmath374 , such that for all possible values of @xmath375 1 .
@xmath376 in @xmath377 .
2 . @xmath378 .
+ we call @xmath379 the _ length _ of the word @xmath380 .
possibly @xmath381 , i.e. @xmath382 for some @xmath383 .
the elements @xmath384 and @xmath385 are called the _ ends _ of the word @xmath380 .
we call an @xmath363-word _ full _ if , whenever @xmath386 is not a unique element in its @xmath320-class , then @xmath387 , and whenever @xmath388 is not a unique element in its @xmath320-class , then @xmath389 .
we denote by @xmath390 the _ inverse word _ @xmath391 and call an @xmath307-word _ symmetric _ if @xmath392 .
we call @xmath380 _ quasisymmetric _ if it can be presented in the form @xmath393 for a shorter word @xmath394 .
we call the end @xmath384 ( @xmath385 ) of the word @xmath380 _ special _ if @xmath395 and @xmath396 ( respectively @xmath397 and @xmath398 ) .
we call the word @xmath380 * _ usual _ if neither of its ends is special ; * _ special _ if one of its ends is special ; * _ bispecial _ if both its ends are special .
+ _ note that a special word is never symmetric , while a bispecial word is always full ; a quasisymmetric word is always bispecial .
if @xmath399 and @xmath400 in @xmath307 , we call the word @xmath401 an _ @xmath363-cycle_. note that in this case @xmath402 is always even . for a cycle @xmath380 we set @xmath403 and @xmath404 for all integers @xmath405 . 6 .
we call an @xmath363-cycle @xmath406 _ non - periodic _ if the sequence @xmath407 can not be written as a multiple @xmath408 of a shorter sequence @xmath409 .
a _ shift _ of a cycle @xmath380 is defined as the cycle @xmath410}=x_{k+1}r_{k+1}x_{k+2}\dots r_{k-1}x_k$ ] for some even integer @xmath411 .
we call a non - periodic cycle @xmath380 _ symmetric _ if @xmath412}$ ] for some @xmath375 .
( note that @xmath410}=w^{[l]}$ ] with @xmath413 is impossible if @xmath380 is non - periodic . )
8 . for a cycle @xmath380 and an even integer @xmath414 we define @xmath415 as the number of even integers @xmath416 such that both @xmath417 and @xmath418 belong either to @xmath419 or to @xmath420 .
[ a13 ] 1 .
a _ usual string _
datum is a non - symmetric full usual word .
2 . a _ special string _
datum is a pair @xmath421 , where @xmath380 is a special full word and @xmath422 .
3 . a _ bispecial string _
datum is a quadruple @xmath423 , where @xmath380 is a bispecial word , which is neither symmetric nor quasisymmetric , @xmath424 and @xmath425 .
4 . a _ band _
datum is a pair @xmath426 , where @xmath380 is a non - periodic cycle and @xmath427 $ ] is a _ primary polynomial _ over the field @xmath146 , i.e. a degree of an irreducible polynomial with leading coefficient @xmath428 , such that @xmath429 and if @xmath380 is symmetric also @xmath430 .
if the field @xmath146 is algebraically closed and @xmath431 , we write @xmath432 instead of @xmath426 . 5 . the following string data are called _ equivalent _ : 1 .
usual string data @xmath380 and @xmath390 ; 2 .
special string data @xmath421 and @xmath433 ; 3 .
bispecial string data @xmath423 and @xmath434 . 6 .
two band data are called _ equivalent _ if they can be obtained from one another by a sequence of the following transformations : 1 .
replace @xmath426 by @xmath435},f)$ ] if @xmath415 is even ; 2 . replace @xmath426 by @xmath435},\al^{-1}t^df(1/t))$ ] , where @xmath436 and @xmath437 if @xmath415 is odd ; 3 .
replace @xmath426 by @xmath438 .
+ _ note that if @xmath439 , then @xmath440 .
_ then the main result of the papers @xcite ( see also @xcite ) can be reformulated as follows .
[ a14 ] there is one - to - one correspondence between isomorphism classes of indecomposable representations of a bunch of chains and equivalence classes of string and band data .
we call indecomposable representations corresponding to usual string data ( special string data , bispecial string data , band data ) _ usual strings _
( respectively , _
special strings , bispecial strings , bands _ ) .
for our purpose we have to consider some infinite representations of a bunch of chains .
we suppose now that @xmath441 and for every index @xmath309 the set @xmath442 is finite .
namely , we define the category @xmath443 just in the same way as @xmath444 , but allowing infinitely many elements of @xmath329 to occur in every representation . on the contrary
, we always suppose that for every @xmath309 the sum of all dimensions @xmath445 with @xmath446 is finite .
the last condition looks indispensable , since even when one considers the simplest case @xmath447 ( which means square matrices under conjugation ) , the classification of representations of infinite dimension is a wild problem .
[ a21 ] let @xmath276 be a locally finite dimensional category , @xmath448 be its full subcategory and @xmath358 be an @xmath276-bimodule .
we say that @xmath358 is _ triangular with respect to @xmath448 _ if , for every indecomposable objects @xmath449 , where @xmath450 and @xmath451 , @xmath452 .
[ a22 ] let @xmath358 be triangular with respect to @xmath448 .
for any object @xmath453 choose a decomposition @xmath454 , where @xmath455 and @xmath456 has no direct summands from @xmath448 .
for a morphism @xmath457 or an element @xmath458 denote , respectively , by @xmath459 or @xmath460 its component from @xmath461 or @xmath462 .
if @xmath457 is a morphism in @xmath444 from @xmath463 to @xmath464 ( i.e. @xmath465 ) , then @xmath459 is a morphism from @xmath460 to @xmath466 .
especially if @xmath88 is an isomorphism @xmath467 , then @xmath459 is an isomorphism @xmath468 .
[ a23 ] let @xmath276 be a locally finite dimensional category that is a union of a chain @xmath469 of full subcategories .
suppose that @xmath358 is an @xmath276-bimodule that is triangular with respect to each @xmath470 .
denote by @xmath471 the category of infinite direct sums @xmath472 , where @xmath473 is an object from @xmath470 with no direct summands from @xmath474 , and by @xmath475 the natural extension of @xmath358 onto @xmath471 .
for each element @xmath476 from @xmath477 denote by @xmath478 its restriction onto @xmath479 . if @xmath480 are two elements such that @xmath481 for all @xmath263 , then @xmath482 . proof .
first suppose the field @xmath146 _ uncountable_. consider the sets of isomorphisms @xmath483 and the natural mappings @xmath484 arising from the triangularity condition .
these sets can be considered as algebraic ( even affine ) varieties , then @xmath485 are open morphisms of these varieties .
in particular , the images @xmath486 form a decreasing chain of non - empty open subsets in @xmath487 .
hence their intersection is also non - empty ( cf .
for instance @xcite ) .
take an element @xmath459 from this intersection and set @xmath488 .
again they are algebraic varieties and @xmath489 are non - empty open subsets of @xmath490 , thus there is an element @xmath491 . continuing this process
, we get a sequence @xmath492 of elements from @xmath483 such that @xmath493 for all @xmath494 .
this sequence defines an isomorphism @xmath495 .
if @xmath146 is arbitrary , take its uncountable extension @xmath496 and consider extensions of @xmath276 and @xmath358 to @xmath496 .
it is easy to see that @xmath482 their extensions are isomorphic , which accomplishes the proof .
[ ] @xmath180 [ a230 ] we use the suppositions and notations of lemma [ a23 ] . if @xmath497 and @xmath498 , denote by @xmath236 the component of @xmath88 belongings to @xmath499 .
the category @xmath500 is @xmath279-local ( in particular , krull
schmidt ) .
moreover , if @xmath501 is an indecomposable element from @xmath500 and @xmath502 , then either all @xmath236 are invertible or all of them are nilpotent .
now we define _ infinite @xmath307-words _ as sequences @xmath503 which are one - side or two - side infinite , subject to conditions ( a ) and ( b ) of definition [ a12](1 ) and such that for each @xmath504 the set @xmath505 is finite .
we apply to such words all terminology from definitions [ a12](24 ) and [ a13](1,2,5ab )
. then we can extend theorem [ a14 ] to infinite representations .
[ a24 ] isomorphism classes of indecomposable infinite representations of a bunch of chains are in with equivalence classes of infinite string data .
moreover , every infinite representation uniquely decomposes into a direct sum of indecomposable ones .
sketch of the proof ( more details will appear in @xcite ) .
let @xmath506 be the bunch of chains with the index set @xmath507 @xmath508 , the same chains @xmath509 and the same relation @xmath314 , @xmath510 .
then we are in the situation of lemma [ a23 ] .
we define representations corresponding to infinite string data just as it has been done in @xcite for finite case .
one can show that all of them are indecomposable and their endomorphism rings are local .
so we only have to prove that there are no more indecomposable infinite representations .
for each representation @xmath511 we denote by @xmath512 the restriction of @xmath365 onto @xmath506 , given by all matrices @xmath368 with @xmath513 .
lemma [ a23 ] implies that @xmath514 @xmath515 for every @xmath379 .
suppose that @xmath365 is infinite and indecomposable and consider an indecomposable direct summand @xmath516 of a representation @xmath512 .
the reduction procedure and the explicit description of strings and bands from @xcite immediately imply the following facts . 1 .
@xmath516 can not be a band or a bispecial string .
if @xmath516 is a usual or a special string , there is an integer @xmath517 and an indecomposable direct summand @xmath518 of @xmath519 such that the word @xmath380 from the string datum corresponding to @xmath516 is a part of the word @xmath520 from the string datum corresponding to @xmath518 .
3 . if @xmath521 is another indecomposable direct summand of @xmath512 , the number @xmath517 and the representation @xmath518 from ( 2 ) can be chosen common for @xmath516 and @xmath521 .
the authors would like to thank the referee for a very careful reading of the paper and numerous remarks .
the first author would like to thank v.m .
bondarenko , + c .- m . ringel and w. srgel for helpful discussions of the results of this article . | in this article we classify indecomposable objects of the derived categories of finitely - generated modules over certain infinite - dimensional algebras .
the considered class of algebras ( which we call nodal algebras ) contains such well - known algebras as the complete ring of a double nodal point @xmath0/(xy)$ ] and the completed path algebra of the gelfand quiver . as a corollary
we obtain a description of the derived category of harish - chandra modules over @xmath1 .
we also give an algorithm , which allows to construct projective resolutions of indecomposable complexes . in the appendix
we prove the krull - schmidt theorem for homotopy categories .
_ key words : _ derived categories ; tame matrix problems ; harish - chandra modules ; krull - schmidt theorem . [ section ] [ theorem]corollary [ theorem]remark [ theorem]example [ theorem]lemma [ theorem]proposition [ theorem]statement [ theorem]definition [ theorem]conjecture [ theorem]problem [ theorem]question # 1\{#1 } # 1#2\{#1|#2 } # 1#1mod # 1#1pro # 1#1 = derived categories of nodal algebras igor burban and yuriy drozd universitt kaiserslautern , fachbereich mathematik , + postfach 3049 , 67653 , kaiserslautern , germany + and + department of mechanics and mathematics , + kyiv taras shevchenko university , 01033 kyiv , ukraine + e - mails : [email protected] , [email protected] |
the absence of preferred frames in local experiments ( or local boost invariance ) is verified everyday in high - energy experiments but is much more difficult to test for the gravitational interaction .
metrically coupled tensor
scalar theories of gravity ( la jordan fierz brans dicke ) , including general relativity , do not predict any violation of this invariance ( see _ e.g. _ @xcite and references therein ) . on the other hand ,
one expects the universe s global matter distribution to select a preferred rest frame for the gravitational interaction if it is mediated in part by a long - range vector field ( or a second tensor field besides the unique one postulated by einstein ) @xcite .
tests of the boost - invariance of gravity in localized systems are therefore of special interest to determine the _ field content _ of the gravitational interaction , _
i.e. _ specifically whether gravity contains , besides the standard tensor contribution and an often considered scalar one , extra contributions due to the exchange of a vector or a second tensor field . in the post - newtonian limit ,
all preferred frame effects are phenomenologically describable by only two parameters , @xmath2 and @xmath3 @xcite .
[ note that the post - newtonian approach assumes that all the fields contributing to gravity are massless , or at least have a range much larger than the size of the considered @xmath4-body system . ]
these two preferred - frame parameters contribute non - boost - invariant terms in the lagrangian , depending on the velocities @xmath5 of the bodies with respect to some gravitationally preferred rest frame .
more precisely , the post - newtonian lagrangian describing the interaction between @xmath4 _ spherical _ bodies reads @xmath6 \nonumber\\ & & -\ , { 1\over 2 } \sum_{b\not= a\not= c } ( 2\beta -1)\ , { g^2m_am_bm_c\over c^2 r_{ab}r_{ac}}\ , \label{eq:2 } \\ l_{\alpha_1 } = & & -\,{\alpha_1 \over 4 } \sum_{a\not= b}\ , { g m_am_b\over c^2 r_{ab}}\ , ( { \bf v}^0_a \cdot { \bf v}^0_b ) \ , \label{eq:3 } \\
l_{\alpha_2 } = & & { \alpha_2\over 4 } \sum_{a\not= b } { gm_a m_b \over c^2 r_{ab}}\ , \left [ ( { \bf v}^0_a \cdot { \bf v}^0_b ) - ( { \bf n}_{ab } \cdot { \bf v}^0_a ) ( { \bf n}_{ab } \cdot { \bf v}^0_b ) \right]\ , \label{eq:4}\end{aligned}\ ] ] where the effective coupling constant @xmath7 for the gravitational interaction of bodies @xmath8 and @xmath9 is @xmath10 \ , \label{eq:4bis}\ ] ] in which @xmath11 denotes the combination of parameters @xmath12 while @xmath13 denotes the gravitational self - energy of body @xmath8 .
apart from the contributions of @xmath2 and @xmath3 to @xmath7 ( _ via _
@xmath11 ) , @xmath14 is the usual boost - invariant lagrangian involving the eddington parameters @xmath15 and @xmath16 , which are sufficient to parametrize the whole class of metrically - coupled tensor multi - scalar theories ( @xmath17 in general relativity , while these parameters can take arbitrary values in tensor scalar theories ) . the parameters @xmath2 and @xmath3 contribute to the effects associated with a violation of the strong equivalence principle ( @xmath18 ) only through the combined parameter @xmath11 .
observational tests of the parameter @xmath11 are discussed in the literature ( _ e.g. _ @xcite ) . in the following we concentrate on the effects associated to the velocity - dependent lagrangian @xmath19 of eq .
( [ eq:3 ] ) .
it has been shown in ref .
@xcite that the close alignment of the sun s spin axis with the solar system s planetary angular momentum yields an extremely tight bound on @xmath3 @xmath20 this limit on @xmath3 is much stronger than the existing limits on the other post - newtonian parameters @xmath15 , @xmath16 and @xmath2 @xcite .
present experimental bounds on the eddington parameters are at the level @xcite @xmath21 while the limits on @xmath2 are about six times tighter @xcite .
more precisely combined orbital data on the planetary system yield @xcite @xmath22 while the bound obtained in the strong - field regime by analyzing binary - pulsar data is comparable @xcite @xmath23 recent theoretical developments in tensor
scalar cosmological models @xcite suggest that a natural level for @xmath24 and @xmath25 is @xmath26@xmath27 .
the generalization of this result to classes of gravitational theories involving extra vector or tensor interactions has not been worked out , but , by analogy , one might expect the present limit on @xmath2 to be too weak to constrain at a significant level the participation to gravity of extra vector or tensor fields .
it is therefore worth discussing experiments having the capability of improving the existing limits on @xmath2 down to the @xmath28@xmath29 level .
the object of the present paper is to show that artificial earth satellites offer very promising tools to improve the precision of measurement of @xmath2 by , possibly , a couple of orders of magnitude .
indeed , we find that the appearance of small divisors can considerably enhance preferred - frame effects when the semi - major axis of the satellite s orbit and/or its inclination with respect to the earth s equatorial plane take particular values .
section ii is devoted to the secular evolution equations satisfied by the orbital elements , whose @xmath2-dependent contributions derive from the disturbing function ( [ eq:10 ] ) . as a first example of a small divisor which enhances preferred - frame effects , we consider in section iii the simple case of equatorial orbits , and show that because of a competition between tidal forces and the quadrupolar moment of the earth , there exists an optimal value of the semi - major axis which maximizes the perturbations due to @xmath2 .
it turns out that geostationary satellites are nearly optimal .
the case of non - equatorial orbits is studied in section iv .
we point out that preferred - frame effects can be enhanced by arbitrarily large factors if the inclination is sufficiently close to one of six specific values , so that the accuracy of measurement of @xmath2 becomes limited only by the finite duration of the experiment . when deriving bounds on @xmath2 or @xmath3 [ such as eqs .
( [ eq:5]),([eq:6]),([eq:7 ] ) ] , it is necessary to make a definite assumption about the preferred rest frame entering the lagrangians ( [ eq:3]),([eq:4 ] ) .
the standard assumption @xcite , that we shall take up in the present paper , is to choose the frame defined by the cosmic microwave background [ this essentially means that the range of the putative extra vector or tensor field responsible for the violation of local boost invariance is at least of cosmological magnitude ] .
however , we shall see in section v.a that ( somewhat weaker ) bounds on @xmath2 can be obtained without the need of such a hypothesis .
indeed , the ` absolute ' velocity of an artificial earth satellite can be decomposed as @xmath30 , where @xmath31 is the velocity of the satellite with respect to the earth , @xmath32 the orbital velocity of the earth around the sun , and @xmath33 the velocity of the sun with respect to the gravitationally preferred frame .
the @xmath2-contribution ( [ eq:3 ] ) to the interaction term between the earth and the satellite reads then @xmath34 this involves a term proportional to @xmath35 which leads to observable effects , although they are 12 times smaller than those proportional to @xmath36 , since @xmath37 whereas @xmath38 @xcite .
section v.b is devoted to the effects involving @xmath39 that are neglected in the rest of the paper .
in particular , we show there that the term proportional to @xmath40 in eq .
( [ eq:8 ] ) leads to sizable deviations of the angular position of the satellite , which could be used to constrain the value of @xmath2 independently of the tests proposed in sections iii and iv .
we finally summarize our results and give our conclusions in section vi .
the preferred - frame effects described by the lagrangians ( [ eq:3 ] ) and ( [ eq:4 ] ) are @xmath41 smaller than the leading newtonian interaction , and lead , over one orbital period , to very small deviations of the position of a satellite . taking into account the present limits ( [ eq:6])-([eq:7 ] ) , one expects @xmath2-induced displacements @xmath42 ( for @xmath43 ) , too small to be observable with present or foreseeable techniques .
fortunately , some of these perturbations build up beyond one orbital period , and can thus be enhanced to an observable level if one waits for a sufficiently large number of periods .
we restrict our attention to such effects in this paper , _
i.e. _ we concentrate on _ secular _ variations of the orbital elements . concentrating on secular effects
has also the advantage of freeing us from the coordinate ambiguities present in orbital - period effects@xcite .
when taking into account only the newtonian potential generated by a spherical earth @xmath44 , the satellite s orbit is determined by six constants of motion : its semi - major axis @xmath45 , its eccentricity @xmath46 , its inclination @xmath47 with respect to the earth s equatorial plane , the angle @xmath48 between a direction of reference @xmath49 @xcite and the ascending node , the angle @xmath50 between this ascending node and the perigee , and finally the quantity @xmath51 entering the mean anomaly @xmath52 , where @xmath53 is the orbital frequency .
[ numerically @xmath54 , when using @xmath55 , @xmath56 . ] to help visualizing the meaning of the angle @xmath51 , it is useful to note that , in the limiting case of a small eccentricity @xmath46 , the mean anomaly @xmath57 can be identified with the angular position of the satellite , so that @xmath51 can be viewed as the angle between the perigee and the satellite at @xmath58 .
figure 1 summarizes this notation .
it will also be useful in the following to define an orthonormal triad @xmath59 , where @xmath60 is directed towards the ascending node and @xmath61 is in the direction of the orbital angular momentum , _ i.e. _ orthogonal to the orbital plane .
( note that this triad is _ not _ the one used in references @xcite , where @xmath60 was directed towards the periastron . )
we find it convenient to use the method of variation of the elements , as described for instance in ref .
@xcite , to derive the secular variations of @xmath45 , @xmath46 , @xmath47 , @xmath51 , @xmath50 and @xmath48 . for more generality ,
let us consider a perturbed two - body system @xmath62 , define @xmath63 , @xmath64 , @xmath65 , and write the lagrangian as @xmath66 where @xmath67 contains , besides the kinetic terms , only the newtonian potential between spherical bodies @xmath68 , and @xmath69 is the disturbing function containing all corrective terms due to asphericities , tidal forces , relativistic effects , we shall not consider here the motion of the center of mass of the two - body system @xcite , but concentrate upon the equations satisfied by the elements of the _ relative _ orbit @xmath70 .
these derive directly from @xmath69 , and read @xcite [ eq:9 ] @xmath71 -{1-e^2\over na^2e}\ { \partial r\over\partial e}\ , \label{eq:9d}\\ { d\omega\over dt } = & & { ( 1-e^2)^{1/2}\over na^2e}\ { \partial r\over\partial e } -{\cot i\over na^2(1-e^2)^{1/2}}\ { \partial r\over\partial i}\ , \label{eq:9e}\\ { d\omega\over dt } = & & { 1\over na^2(1-e^2)^{1/2}\sin i}\ { \partial r\over\partial i}\ \cdot \label{eq:9f}\end{aligned}\ ] ] the brackets in the right hand side of eq .
( [ eq:9d ] ) indicate that the @xmath45-differentiation is effected keeping the mean anomaly fixed ( _ i.e. _ ignoring the implicit @xmath45-dependence contained in @xmath72 ) .
when working to first order in @xmath69 , the orbital elements can be replaced by their constant zeroth order values on the right - hand - sides
. then the secular variations of the orbital elements are simply obtained by replacing in equations ( [ eq:9 ] ) @xmath69 by its average value @xmath73 , computed over one unperturbed newtonian orbit [ more precisely , one takes the average over the angle @xmath57 , keeping the other elements fixed ] .
it is straightforward to integrate the terms proportional to @xmath74 in the two - body versions of the lagrangians ( [ eq:2 ] ) and ( [ eq:3 ] ) . decomposing the ` absolute ' velocities as @xmath75 , where @xmath76 is the velocity of the center - of - mass of the system @xmath77 with respect to the preferred rest frame
, we find the contribution @xmath78\over 1+(1-e^2)^{1/2}}\nonumber\\ & & + \left({gm\over ac^2}\right)^2c^2\left({2\gamma-\beta+2+\alpha_1x_a x_b\over ( 1-e^2)^{1/2}}-{8\gamma+7-x_a x_b(1 - 4\alpha_1)\over 8}\right),\label{eq:10}\end{aligned}\ ] ] where the square brackets denote the mixed product @xmath79 , the vector @xmath80 being the lagrange - laplace ( -runge - lenz ) eccentricity vector with norm @xmath46 , directed towards the periastron of body @xmath8 .
the first line comes from the term proportional to @xmath81 in @xmath19 ( [ eq:3 ] ) , and the second from the term involving @xmath82 .
the third line is responsible for the usual relativistic perigee advance and comes from the @xmath74 terms of the lagrangian ( [ eq:1 ] ) , with extra contributions due to the @xmath83 term of @xmath19 .
we have not included in eq .
( [ eq:10 ] ) the contribution of @xmath3 , relying on the limit ( [ eq:5 ] ) to consider that this type of preferred effects is already plausibly excluded . up
to here , we have considered a two - body system with arbitrary masses @xmath62 for generality s sake . in the particular case of artificial earth satellites ,
one can however neglect @xmath84 with respect to @xmath85 , and therefore write @xmath86 , @xmath87 , @xmath88 .
hence , we see from the last line in eq .
( [ eq:10 ] ) that the additional contribution to the relativistic perigee advance generated by @xmath2 is completely negligible . as for the first term on the right hand side of eq .
( [ eq:10 ] ) , it only contributes to the evolution of the element @xmath51 and will be studied in section v.b . in the next two sections we concentrate on the evolution of the other elements , generated by the second term on the r.h.s .
( [ eq:10 ] ) . the contribution ( [ eq:10 ] ) should be added to the usual newtonian multipolar and tidal perturbations , notably the effect of the earth s quadrupolar moment ( @xmath89 ) and the tidal forces due to the moon ( denoted as @xmath90 ) and the sun ( denoted as @xmath91 ) , which are the dominant ones . for analytical simplicity
, we shall not consider any other newtonian perturbation in the present paper .
our estimates of the measurability of @xmath2 will correspondingly be only indicative . to get reliable estimates of the bounds on @xmath2 which could be experimentally obtained
, it would be necessary to perform detailed numerical simulations taking into account all known newtonian and general relativistic effects , as well as all modelizable sources of noise .
[ there exist in particular resonances between several multipolar moments of the earth which have significant effects on long times . ]
the secular contribution due to the earth s oblateness can be written down very easily @xmath92 the contribution of tidal forces is more complicated ( even if one neglects the eccentricities of the earth s and the moon s orbits , as well as their inclinations with respect to the earth s equatorial plane ) because it involves explicitly the angular positions of the moon and the sun .
however , it can be simplified very much if the variations of the orbital elements ( [ eq:9 ] ) are studied on longer time scales , more precisely if one averages over the orbital periods of the moon around the earth ( @xmath93 one month ) and of the earth around the sun ( one year ) .
the variations of the satellite s orbital elements after such averagings can then be derived from @xmath94\ , \label{eq:14}\ ] ] where @xmath95 [ the notation @xmath4 is reminiscent of the fact that @xmath96 is the orbital frequency of the earth around the sun ; note however that @xmath97 is _ not _ the orbital frequency of the moon around the earth , which would involve @xmath98 instead of @xmath99 . ]
beware that the above expression ( [ eq:14 ] ) should not be used to derive variations with periods @xmath100 one year .
the secular variations of the satellite s orbital elements can now be easily derived from eqs .
( [ eq:9 ] ) .
they tell us that the semi - major axis undergoes no secular variation ( @xmath101 ) , and that the change of the inclination is negligible in the limit of a small eccentricity [ @xmath102 .
we shall study the equations satisfied by @xmath46 , @xmath50 and @xmath48 in sections iii and iv , and see that they can be rewritten more compactly as a vectorial equation for the eccentricity vector @xmath80 in the limit @xmath103 .
section v.b will be devoted to the equation satisfied by @xmath51 .
finally , it should be noted that in the main body of the paper we always deal with the motion of a satellite as seen in a ( locally inertial ) geocentric frame .
the @xmath2-dependent effects linked to the connection between such a geocentric frame and a global , barycentric one will be briefly discussed in the concluding section .
the aim of this section is to show on the simple example of equatorial orbits how small divisors can enhance the preferred - frame effects associated to @xmath2 . [
as said above , we neglect the eccentricities of the earth s and the moon s orbits , as well as their inclinations with respect to the earth s equatorial plane . ]
the secular variations of the satellite s orbital elements are then given by equations ( [ eq:9 ] ) , where @xmath73 is the sum of eqs . ( [ eq:10 ] ) , ( [ eq:11 ] ) and ( [ eq:14 ] ) with @xmath104 .
however , the angular position of the ascending node is no longer well - defined as @xmath105 , and equations ( [ eq:9e ] ) and ( [ eq:9f ] ) become formally singular .
as is well known @xcite , this singularity is fictitious and taken care of by considering the evolution of the angle @xmath106 between the direction of reference @xmath49 and the perigee , which stays well - defined in the limit @xmath105 .
one finds @xmath107\over e^2 } + o(\alpha_1 e ) \ , \label{eq:15}\end{aligned}\ ] ] where @xmath108 \label{eq:16}\ ] ] is the newtonian perigee advance due to the earth s quadrupolar moment and tidal forces , and @xmath109 the velocity of the earth with respect to the preferred frame , @xmath32 denoting as in section i the orbital velocity of the earth around the sun .
[ it is useful to quote here the numerical value of @xmath110 , which will appear again in the following sections as the characteristic frequency for the newtonian variations of the orbital elements @xmath50 and @xmath48 ; see notably eqs .
( [ eq:26 ] ) and ( [ eq:28 ] ) . ] equation ( [ eq:15 ] ) together with the one satisfied by @xmath46 @xmath111 can be rewritten as a simple vectorial equation for the eccentricity vector @xmath80 in the limit @xmath103 @xmath112 + o(\alpha_1 e^2)+o(e^2 ) \ , \label{eq:18}\ ] ] where @xmath113 is a constant vector ( beware that ref .
@xcite defines a vector @xmath114 which is equal to twice this value ) and @xmath115 a yearly varying one . in the present and
following sections , we concentrate on the effects of the constant vector @xmath114 , leaving to section v.a a study of the effects associated to @xmath116 .
since we are considering equatorial orbits in this section , the vector @xmath117 orthogonal to the orbital plane is the constant unitary vector parallel to the earth s polar axis , and therefore @xmath118 is a _ constant _ vector , namely the projection of @xmath114 onto the equatorial plane .
hence a @xmath2-type violation of local boost invariance has the consequence of adding a constant forcing term in the time evolution of the eccentricity vector which , if it were alone , would secularly ` polarize ' the orbit in the direction of the equatorial projection of @xmath33 .
however , the newtonian precession term @xmath119 cuts off the build up of this polarizing term and deflects it by @xmath120 in a gyroscope - like way .
more precisely , the solution of eq .
( [ eq:18 ] ) can be written as the vectorial superposition @xmath121 where @xmath122 is a vector of constant norm rotating with angular frequency @xmath119 in the orbital plane ( usual newtonian perigee advance ) , and @xmath123 is the fixed polarizing contribution due to the preferred - frame effects we are studying , namely @xmath124 note the factor @xmath125 in eq .
( [ eq:21 ] ) which is the first appearance of a small divisor .
the solution ( [ eq:20 ] ) is formally identical to the one found for binary pulsars in ref .
@xcite , the only difference being that @xmath119 replaces the relativistic periastron advance .
the same kind of polarizing terms in the time evolution of @xmath80 has also been pointed out previously in @xcite , in the totally different context of equivalence principle violation in binary systems . in geometrical terms , eq .
( [ eq:20 ] ) means that the eccentricity vector @xmath126 traces out , during its time evolution ( after averaging over an orbital period ) , a circle centered around @xmath127 .
if a large enough segment of this circle ( say about a quarter ) can be monitored during the experiment , it should be possible to measure the position of its center , _
i.e. _ @xmath127 , with about the same precision that individual measurements of the eccentricity vector @xcite . in turn
, the precision of the measurement of @xmath80 is related to the precision @xmath128 with which one can measure the satellite s position : roughly one expects @xmath129 .
it is convenient to work with quantities homogeneous to displacements .
therefore we shall measure the @xmath2-perturbations of @xmath80 in terms of @xmath130 which represents ( when @xmath131 , as expected from @xmath132 ) the amplitude of the secular change in the distance to the perigee @xmath133 . from the arguments
just given , equating @xmath134 with the position measurement precision @xmath135 should yield an estimate of the precision with which @xmath2 can be measured over a time span @xmath136 , where @xmath137 the denominator of the expression ( [ eq:22 ] ) is a function of the semi - major axis @xmath45 , and is minimized for the value @xmath138 ( which corresponds to an orbital period @xmath139 hours ) . in eq .
( [ eq:23 ] ) , we have used the numerical data @xmath140 , @xmath141 , @xmath142 , and @xmath143 . the maximal value of @xmath134 ( for equatorial orbits ) is thus obtained for this value of the semi - major axis , and reads @xmath144 where the norm @xmath145 and the direction @xmath146 of @xmath33 have been extracted from the results of cobe @xcite .
present technologies make it possible to measure the position of a satellite down to @xmath147 , by using either laser ranging ( as for the laser geodynamical satellites ` lageos ' ) or global positioning system ( gps ) receivers onboard .
the displacement ( [ eq:24 ] ) could therefore lead in principle to limits on @xmath2 of the order of @xmath148 , _
i.e. _ two orders of magnitude tighter than the present bounds ( [ eq:6 ] ) and ( [ eq:7 ] ) .
it is to be noted that a geostationary satellite ( @xmath149 , @xmath150 ) is near optimal .
it yields @xmath151 , leading us to expect a precision @xmath152 .
clearly , numerical simulations taking into account all known newtonian perturbations and modelizable non - conservative forces should be performed in order to get more realistic estimates of the precision which can be reached .
this simple case of equatorial orbits already exhibits one of the characteristic behaviors of preferred - frame effects enhanced by small divisors : the duration of the experiment must be large enough to take advantage of the small divisor effect .
for instance , the time span of the experiment should be ( at least ) @xmath153 in the case of the optimal orbit ( [ eq:23 ] ) , and @xmath154 for a geostationary satellite .
let us conclude this section by a comparison of the result ( [ eq:24 ] ) with the one corresponding to the best drag - free satellite we know : the moon itself . still neglecting the inclination of its orbit and that of the ecliptic with respect to the earth s equatorial plane , we can repeat all of the above discussion , except that in eq .
( [ eq:13 ] ) only the tidal forces due to the sun should be taken into account .
actually , higher - order terms in these tidal forces are not small in the particular case of the moon ( because its orbital period is non - negligible as compared to one year ) , and the coefficient @xmath155 of eq .
( [ eq:14 ] ) should be replaced by @xmath156 instead of merely @xmath157 to yield the correct magnitude of @xmath119 @xcite .
the amplitude of the secular oscillation of the moon s perigee distance is then given by the same formula ( [ eq:22 ] ) as above with @xmath158 .
the earth s oblateness gives then a negligible contribution to the denominator of @xmath159 in eq .
( [ eq:22 ] ) , and one finds @xmath160 , with period @xmath161 . this effect was first discussed ( in a different guise ) in ref .
since lunar laser ranging ( llr ) allows one to determine the earth moon distance within 2 or 3 centimeters , we expect that an analysis of llr data could at best limit @xmath2 at the @xmath162 level , _
i.e. _ not better than the existing bounds ( [ eq:6]),([eq:7 ] ) .
however , as this would represent a new , independent test of preferred - frame effects , it would be interesting to perform explicitly a multi - parameter fit of llr data including the contributions of @xmath2 ( and for completeness @xmath3 , @xmath15 , @xmath16 ) to the lunar equations of motion .
in this section , we shall see that non - equatorial orbits ( of any altitude ) can give rise to arbitrarily small divisors if the inclination is near some special values . for reasons that will appear clear below , we concentrate upon rather low satellites , for which tidal forces are negligible compared to the influence of the earth s quadrupolar moment . for this reason we do not take into account the disturbing function ( [ eq:14 ] ) in the present section , in order to avoid unnecessarily technical calculations . however , we show at the end of this section how the equations satisfied by the orbital elements can be solved when tidal forces are not neglected .
the main difference between equatorial and non - equatorial satellites is that the orbital plane is no longer fixed when @xmath163 . indeed ,
equation ( [ eq:9f ] ) shows that it is precessing with a constant angular frequency @xmath164 which is not modified by a @xmath2-type violation of local lorentz invariance in the limiting case of a small eccentricity
. we shall drop the angular brackets in the following , and denote this constant precessing velocity simply by @xmath165 .
like in the previous section , the equations ( [ eq:9b ] ) and ( [ eq:9e ] ) satisfied by @xmath46 and @xmath50 can be rewritten as a simple vectorial equation for the evolution of the components of @xmath80 with respect to the vectors @xmath166 , which are part of a frame [ defined at the beginning of section ii and in fig
. 1 ] which rotates around the earth s polar axis with the angular velocity ( [ eq:26 ] ) : @xmath167 + o(\alpha_1 e^2)+o(e^2)+o(en^2 ) \ , \label{eq:27}\ ] ] where the prime in @xmath168 denotes a time derivative in the rotating frame @xmath169 and where @xmath170 is the newtonian perigee advance due to the earth s quadrupolar moment , @xmath171 and @xmath172 being the vectors defined in eqs .
( [ eq:19]),([eq:19bis ] ) . as in section iii
, we shall neglect in this section @xmath172 with respect to @xmath114 . since the projection @xmath118 of @xmath114 onto the orbital plane is no longer a constant vector because of the precession @xmath165 , the solution ( [ eq:20 ] ) of the previous section is not valid for non - equatorial orbits .
however , it is easy to solve equation ( [ eq:27 ] ) which is just an inhomogeneous _ linear _ differential equation in @xmath80 ( in the limiting case of a small eccentricity ) .
let us look for a solution of the type @xmath173 , where @xmath174 is the usual ( constant - norm ) newtonian eccentricity vector rotating in the orbital plane @xmath166 with angular frequency @xmath175 , and where the @xmath176 s are some constant - norm vectors rotating in the orbital plane with constant angular frequencies @xmath177 to be determined .
the time derivative of @xmath80 then reads @xmath178 this has precisely the form of equation ( [ eq:27 ] ) , with @xmath179 having been decomposed as a sum of constant - norm vectors @xmath180 rotating with constant angular frequencies @xmath177 in the orbital plane . since @xmath181 has a periodic motion in the orbital plane ( with period @xmath182 ) , it obviously admits such a decomposition : @xmath183 where [ eq:31 ] @xmath184 \ , \label{eq:31a } \\ { \bf k}_0\equiv & & \sin\delta\ , \sin i\ { \bf a } \ , \label{eq:31b}\end{aligned}\ ] ] @xmath185 denoting the right - ascension and declination of @xmath33 .
@xmath186 are rotating with angular frequencies @xmath187 in the orbital plane , whereas @xmath188 is a constant vector in this plane ( directed towards the ascending node ) .
the solution of eq .
( [ eq:27 ] ) can therefore be written simply as @xmath189 this generalizes the solution ( [ eq:20 ] ) obtained above for equatorial orbits .
[ in the limit @xmath105 , the term involving @xmath190 becomes equal to the vector @xmath123 of eq .
( [ eq:21 ] ) , in which @xmath191 was denoted @xmath119 . ] in geometrical terms , eq .
( [ eq:32 ] ) means that the eccentricity vector @xmath126 undergoes a kind of epicyclic motion in the @xmath166 plane : it moves ( with angular velocity @xmath175 ) along a circle of radius @xmath192 whose center moves itself ( with a non - uniform angular velocity determined by @xmath165 ) on an ellipse @xmath193 centered around @xmath194 .
hence we can now distinguish three different contributions to the @xmath2-induced secular oscillation of the distance to the perigee : [ eq:33 ] @xmath195 where @xmath196 figure 2 displays the amplitude of these displacements as functions of the inclination .
they clearly exhibit poles ( or ` resonances ' ) for six particular values of @xmath47 : @xmath197 diverges for @xmath198 , _
i.e. _ for @xmath199 or @xmath200 ; @xmath201 diverges for @xmath202 , _
i.e. _ for @xmath203 or @xmath204 ; and @xmath205 diverges for @xmath206 , _
i.e. _ for @xmath207 or @xmath208 .
( note that the special value @xmath209 coincides with the well - known newtonian critical inclination for @xmath89-effects @xcite . ) as in the equatorial case discussed above , the price to pay for taking advantage of the small divisors arising near these poles is the need for a long observation period , inversely proportional to the corresponding small divisor .
indeed , the dephasing periods between the newtonian eccentricity vector @xmath174 and the contributions @xmath210,@xmath194 proportional to @xmath2 read [ eq:35 ] @xmath211 where @xmath212 [ near each resonance @xmath213 , one of the vectors @xmath210 becomes infinitely large with respect to the other , and the @xmath214 ellipse degenerates to a circle described with angular velocity @xmath215 . ] in order to compare the merits of the different resonances , it is useful to define as a figure of merit the ratio of the corresponding perigee displacement @xmath159 by the typical ( minimal ) time @xmath136 needed for the observation : [ eq:37 ] @xmath216 these results underline that the inclinations around @xmath217 or @xmath208 have the capability to give the largest effects in a given observational time , as confirmed by the width of the resonances in fig .
2 . equations ( [ eq:37 ] ) also show that low orbits ( say @xmath218 ) seem preferable in that they give larger @xmath159 s in a given observational time .
[ it is easy to see that the @xmath219 dependence of @xmath220 still applies for high , tidally perturbed orbits .
this justifies our concentrating on low orbits , with negligible tidal effects .
] however , equations ( [ eq:35]),([eq:36 ] ) show that the inclination @xmath47 must be fixed with a high precision , typically within less than one arc - minute if one wishes to make the fullest use of the observational time @xmath136 which , for practical reasons , will probably not exceed @xmath221 years .
the needed inclination can be computed thanks to the following asymptotic formulae : [ eq:38 ] @xmath222 for instance a low satellite ( @xmath223 ) will give rise to a perigee displacement @xmath224 in @xmath225 if the inclination differs from the pole value @xmath226 ( or @xmath208 ) by only @xmath227 .
an error of one arc - minute on @xmath47 would change the observational time and the perigee displacement by factors @xmath228 .
the six resonant values of the inclination are solutions of the simple trigonometric equations @xmath229 , where @xmath230 , and do not depend on any experimental data .
they will therefore enhance preferred - frame effects in the motion of the natural satellites of the different planets in the solar system ( although tidal forces may not be negligible is some cases ) .
however , none of the known natural satellites orbits has an inclination close enough to one of the six poles . on the other hand
, there are hundreds of artificial earth satellites , and many of them would allow one to tighten the present bounds ( [ eq:6]),([eq:7 ] ) on @xmath2 if the evolution of their perigee could be tracked at the centimeter level .
for instance , the classes of satellites ` gps bii ' and ` goes ' would typically allow one to measure perigee radial displacements of @xmath231 in an observational time of @xmath232 .
even if these satellites could be tracked with sufficient precision , it would remain to see whether the effect of non - gravitational forces would allow one to make full use of such long data span .
however , a statistical study of these @xmath233 satellites could probably allow a significant reduction of the sources of errors .
the same remarks can be formulated about the classes of satellites ` gps ' and ` glonass ' whose perigees are deviated by @xmath234 in @xmath235 .
[ there are also many low ( _ i.e. _ fast ) satellites , like ` starlette ' or ` seasat ' , which give perigee oscillations of @xmath236 in less than one year , but the air drag is very large for such low satellites . ]
the first laser geodynamical satellite ( lageos i ) is an interesting candidate , not only because its position has been laser tracked for years at the few centimeter level and because it is submitted to very small air drag , but also because _ several _ contributions to the eccentricity vector @xmath80 are large in its case : @xmath237 gives a perigee deviation of @xmath238 in @xmath239 years , @xmath240 a deviation of @xmath241 in @xmath242 months , and @xmath194 a deviation of @xmath243 in @xmath244 years .
the superposition of these three effects thus yields a complex signal which should , hopefully , be distinguishable from other , newtonian , contributions . unfortunately , the variations of the eccentricity @xmath46 of this satellite are not very precisely modelizable , as explained in refs .
@xcite : it undergoes small oscillations of @xmath245 , which correspond to perigee deviations of @xmath246 .
the recently launched second lageos , which has an even more favorable inclination , might provide another interesting experimental probe of preferred - frame effects .
indeed , a duration of only @xmath247 would suffice to observe a radial displacement of its perigee by @xmath248 .
the best tool for tightening the present bounds ( [ eq:6]),([eq:7 ] ) on @xmath2 would be a ( naturally or artificially ) drag - free satellite , launched on a favorably inclined orbit and tracked for years at the centimeter level ( _ via _ laser ranging or onboard gps receivers ) .
let us note that there are plans for launching in the near future drag - free satellites with gps receivers onboard : gravity probe b ( gpb ) and the satellite test of the equivalence principle ( step ) . for completeness
, we now briefly discuss how the above solution ( [ eq:32 ] ) for the eccentricity vector @xmath80 is modified when tidal forces ( which are very small for low satellites ) are taken into account .
the first ( trivial ) modification is of course that the disturbing function ( [ eq:14 ] ) now gives a contribution to the precession velocity @xmath249 ( [ eq:9f ] ) of the orbital plane : @xmath250+o(\alpha_1 e)+o(e^2)\ . \label{eq:39}\ ] ] on the other hand , the effect of tidal forces in equations ( [ eq:9b ] ) and ( [ eq:9e ] ) for @xmath46 and @xmath50 is somewhat more involved [ even after averaging over monthly and yearly frequencies as explained in section ii , eq . ( [ eq:14 ] ) ] .
the vectorial equation satisfied by the eccentricity vector @xmath80 in the orbital plane reads now @xmath251 where @xmath252 \label{eq:41}\ ] ] is the analog of the angular frequency @xmath175 of eq .
( [ eq:27 ] ) , and @xmath253 is an additional contribution due to tidal forces .
the ratio @xmath254 can become positive and larger than unity when the inclination @xmath47 is very close to the well - known @xmath89-critical inclination @xmath255 or @xmath204 ( _ e.g. _
@xmath256 for @xmath257 , and @xmath258 for @xmath259 ) . equation ( [ eq:40 ] )
then formally exhibits an exponential blow up of the newtonian eccentricity vector .
this indicates that our simplified ( linearized and time - averaged ) treatment of the evolution of @xmath80 becomes inadequate . in the following ,
we restrict our attention to the generic case @xmath260 .
one finds that @xmath261 has no longer a constant norm , as opposed to eq .
( [ eq:20 ] ) and ( [ eq:32 ] ) above , but that it is now moving on an ellipse , whose axes are directed along @xmath60 and @xmath262 and have a ratio @xmath263 : @xmath264 [ note that this ellipse reduces to a circle in the particular case of equatorial orbits ( @xmath104 ) that we considered in section ii . ]
the solution of equation ( [ eq:40 ] ) can then be obtained in a geometrical manner similar to the one used in eqs .
( [ eq:29])([eq:32 ] ) .
let @xmath176 denote a generalization of the newtonian eccentricity vector @xmath174 when the frequency @xmath265 is replaced by an arbitrary constant @xmath266 such that @xmath267 , _
i.e. _ @xmath268 or @xmath269 ( we shall see below that these values of @xmath266 are sufficient for our purpose ) . like @xmath174 in eq .
( [ eq:43 ] ) above , @xmath176 is moving on an ellipse in the orbital plane , with a non - uniform velocity which has the sign of @xmath266 , and such that its components along @xmath60 and @xmath262 are oscillating with a constant frequency @xmath270 and have an amplitude ratio @xmath271 .
[ the particular cases of constant vectors along @xmath60 or @xmath262 are obtained respectively for @xmath272 and @xmath273 .
the time derivative of a linear combination @xmath274 then reads @xmath275 this has precisely the form of eq .
( [ eq:40 ] ) , with @xmath179 having been decomposed as a sum of vectors @xmath276 moving on ellipses as described above .
it is easy to check that any vector having a periodic motion in the orbital plane admits such a decomposition , and that the corresponding @xmath266 s always satisfy the condition @xmath277 .
in particular , one finds that the source term @xmath278 can be decomposed as in eq .
( [ eq:30 ] ) , where @xmath188 is still given by eq .
( [ eq:31b ] ) , but where @xmath186 have more complicated expressions @xmath279 \ , \label{eq:44bis}\ ] ] with @xmath280 note that these expressions for @xmath186 reduce to those of eq .
( [ eq:31a ] ) when @xmath281 , _
i.e. _ when tidal forces are neglected or for equatorial orbits
. hence the solution of eq .
( [ eq:40 ] ) can be written simply as @xmath282 it is therefore of the same kind as eq .
( [ eq:32 ] ) , and there still exist poles for some particular values of the inclination , only slightly modified with respect to those of fig . 2 for relatively low satellites .
[ for the same reason indicated above , our treatment becomes inadequate when @xmath283 , formally corresponding to @xmath284 becoming large and pure imaginary . ]
in the previous sections , we have neglected the orbital velocity @xmath32 of the earth around the sun , but we shall see below that it can also lead to significant preferred - frame effects on artificial satellites , although they will typically be 12 ( @xmath285 ) times smaller than those proportional to @xmath286 .
the interest of effects involving @xmath32 is twofold .
first of all , we shall see that arbitrarily small divisors can enhance the preferred - frame effects we are studying for _ any _ value of the inclination @xmath47 of the satellite s orbit ( if the semi - major axis @xmath45 is chosen appropriately ) instead of the discrete resonances displayed in fig . 2 .
moreover , the existence of a motion around the sun ( with well - defined amplitude and phase ) is known for sure to be part of the ` absolute ' velocities entering preferred - frame effects . by contrast , the identification of @xmath33 with our velocity with respect to the cosmic microwave background is a specific assumption .
although this assumption is plausible on field - theoretical grounds , the bounds derived on @xmath2 in the literature @xcite and the discussions of the previous sections of the present paper are strongly dependent on it .
it seems therefore of importance to determine limits on @xmath2 by relying only on an unambiguously present velocity such as @xmath32 .
it is straightforward to generalize the results of the previous section to the case of a velocity @xmath32 which is not constant , as opposed to @xmath33 .
the equation satisfied by the eccentricity vector @xmath80 is eq .
( [ eq:27 ] ) , but we concentrate now on the effects generated by the source term @xmath287 , where @xmath172 , given by eq .
( [ eq:19bis ] ) , is a vector rotating with angular frequency @xmath288 .
as in the previous section , it suffices to decompose @xmath289 as a sum of constant - norm vectors rotating with constant angular frequencies to derive the contributions to @xmath80 involving @xmath290 .
let us denote as @xmath291 the inclination of the ecliptic with respect to the earth s equatorial plane , and choose the origin of time at the vernal equinox .
the source term @xmath289 can then easily be decomposed as @xmath292 where @xmath293 \ , \label{eq:47}\ ] ] and if @xmath294 , @xmath295 denote two independent signs ( @xmath296 , @xmath297 ) @xmath298 \ . \label{eq:48}\ ] ] the vectors @xmath299 are rotating with angular velocities @xmath300 in the orbital plane , whereas @xmath301 are rotating with angular velocities @xmath302 in this plane .
these notations are chosen to simplify the expression of the eccentricity vector @xmath80 written below ; it generalizes the notations @xmath186,@xmath188 introduced in section iv , which would be denoted @xmath303,@xmath304 in the present convention .
the eccentricity vector can then be immediately written as @xmath305 to which should be added the contributions proportional to @xmath286 of eq .
( [ eq:32 ] ) . the solution ( [ eq:49 ] )
now exhibits resonances for any value of the inclination @xmath47 provided the semi - major axis @xmath45 is chosen appropriately .
figure 3 displays the values of @xmath45 and @xmath47 for which one of the divisors in eq .
( [ eq:49 ] ) vanishes .
note that when @xmath45 tends formally towards @xmath306 , these curves tend towards the six poles found in section iv , because @xmath175 and @xmath165 [ eqs .
( [ eq:28 ] ) and ( [ eq:26 ] ) ] are proportional to @xmath307 whereas @xmath308 is constant . as in the previous section ,
the price to pay for a small divisor is the need for a correspondingly long time of observation , say @xmath309 , or @xmath310 , depending upon the concerned divisor .
the figure of merit @xmath311 of the different orbits can be computed like in eqs .
( [ eq:37 ] ) , and figure 4 shows that the contributions @xmath312 and @xmath313 give the best results , _
i.e. _ the largest perigee displacements in a given observational time .
this is due to the rather small value of the inclination @xmath314 of the ecliptic with respect to the equatorial plane , since @xmath312 and @xmath313 are precisely the only contributions which do not vanish as @xmath315 .
the values of @xmath45 and @xmath47 for which these two main contributions diverge have been plotted in solid lines in fig . 3 in order to distinguish them from those involving @xmath316 or @xmath317 ( vanishing when @xmath318 ) which have been plotted in dashed lines . the largest figure of merit displayed in fig . 4
is obtained for @xmath223 and @xmath319 , and reads @xmath320 . as expected , it is ( @xmath321 ) 12 times smaller than the best result ( [ eq:37a ] ) of section iv , which was obtained for @xmath199 ( or @xmath208 ) .
however , the high - precision observation of a low orbit with an inclination close to @xmath322 could lead to the first experimental results about preferred - frame effects free from any assumption about the fundamental rest frame .
[ the satellite lace happens to be precisely on such an orbit , but it is not drag - free . ] we have seen in section iv that all of the three contributions @xmath323 , @xmath240 and @xmath194 ( proportional to @xmath286 ) to the eccentricity vector @xmath80 are _ a priori _
large for the orbit of lageos i , thus giving rise to a complex signal in the evolution of the eccentricity vector .
it is interesting to quote that the contribution @xmath313 ( proportional to @xmath290 ) is also enhanced by a small divisor for this satellite , since the amplitude @xmath324 is about @xmath325 in a typical observational time of @xmath326 months .
it would be therefore interesting to analyze the lageos i data to look for preferred - frame signals of both the @xmath33 and the @xmath327 types [ if it turns out feasible to extract such information in presence of the non - modelizable forces that act on the satellite ] . a remarkable feature of fig .
3 is the existence of maxima in the curves representing the loci of the @xmath32-type resonances in the @xmath328 plane .
these maxima occur at @xmath104 or @xmath329 ( for @xmath330 , @xmath331 and @xmath332 ) , @xmath333 or @xmath334 ( for @xmath335 ) , and @xmath336 ( for @xmath337 ) .
the amplitude of the @xmath327-preferred - frame effects for orbits located close to these values of @xmath47 and @xmath45 would be therefore almost insensitive to injection errors or fluctuations in @xmath47 .
this suggests that these orbits might be especially robust tools for constraining this type of effects .
the best choices would be the orbits located close to ( @xmath104 , @xmath338 ) or ( @xmath339 , @xmath340 ) , which correspond to rather large figures of merit @xmath341 in fig .
4 ( respectively @xmath342 and @xmath343 ) .
we have not yet discussed the perturbation of the element @xmath51 , which is related to the angular position of the satellite at @xmath58 . using the expression of the disturbing function ( [ eq:10 ] ) in ( [ eq:9d ] ) , the secular evolution of @xmath51 is given by @xmath344\over e^2 } - { \alpha_1 n\over c^2}\ , ( { \bf w}+{\bf v}_\oplus)^2 + o(\alpha_1 e ) \ , \label{eq:50}\ ] ] where @xmath345 is the newtonian contribution due principally to tidal forces ( [ eq:14 ] ) and to the earth s quadrupolar moment ( [ eq:11 ] ) .
the second term of the right - hand - side , though proportional to @xmath346 , does not lead to any interesting effect in the limit of a small eccentricity .
indeed , it precisely cancels a corresponding term in the secular evolution of @xmath50 . in other words , the angular position of the satellite with respect to the ascending node is not affected by this type of contribution .
by contrast , the last term proportional to @xmath347 induces interesting effects on the position of the satellite through the time variation of @xmath40 . [ the secular drift in @xmath51 due to
the constant term @xmath348 is unobservable because , as we shall see below , it can be absorbed in a small renormalization of newton s constant @xmath349 . ] indeed , let us denote as before the right ascension and declination of @xmath33 as @xmath185 , the orbital frequency of the earth as @xmath350 , and the inclination of the ecliptic as @xmath351 . equation ( [ eq:50 ] ) can then be rewritten as @xmath352 \ , \label{eq:51}\ ] ] where we have written down only the largest contribution .
the integration of this equation shows that the longitude of the satellite is modulated by an oscillating term @xmath353 corresponding to an along - track oscillation @xmath354 the large amplitude , yearly periodicity , and weak altitude dependence ( @xmath355 ) of this effect makes it a promising way of improving the precision of measurement of @xmath2 ( maybe down to the @xmath29 level ) .
however , detailed numerical simulations are needed to assert whether the effect ( [ eq:53 ] ) can be separated from the other yearly perturbations .
let us note finally that the results ( [ eq:52]),([eq:53 ] ) can be derived in a totally different way , by starting directly from the lagrangians ( [ eq:1 ] ) and ( [ eq:8 ] ) .
the terms proportional to @xmath2 can indeed be interpreted as a time - dependent renormalization of the gravitational constant @xcite experienced by the satellite @xmath356 \ . \label{eq:54}\ ] ] taking into account the adiabatic constancy of the delaunay variable ( or action variable " ) @xmath357^{1/2}$ ] , eq .
( [ eq:54 ] ) induces a variation of the semi - major axis @xmath358^{-1 } \ . \label{eq:55}\ ] ] in the limiting case of small eccentricity , the angular position @xmath359 of the satellite with respect to the perigee can be identified with the mean anomaly @xmath360 , where @xmath361 is a constant and where the orbital frequency @xmath362 reads @xmath363 ^ 2 \ . \label{eq:56}\ ] ] integrating eq .
( [ eq:56 ] ) reproduces the result ( [ eq:52 ] ) .
note that here again a very small instantaneous perturbation @xmath364 has been enhanced by a large factor @xmath365 . as an aside , let us remark that another consequence of the time - dependent renormalization ( [ eq:54 ] ) of the gravitational constant is to cause a yearly breathing " of the radius of the earth , with associated yearly variations of its moment of inertia and of its angular velocity ( see @xcite ) .
however , the amplitude of these variations , given the present limits on @xmath2 , are too small to be of observational significance , _
e.g. _ @xmath366 .
artificial earth satellites can be very useful tools to probe the field content of gravity , _ i.e. _ specifically whether it contains a vector or second tensor interaction leading to preferred - frame effects in local gravitating systems .
thanks to the appearance of small divisors which enhance the preferred - frame effects on the eccentricity vector when the inclination and/or the semi - major axis of an orbit are chosen appropriately , it seems conceivable to tighten the present experimental bounds on the preferred - frame parameter @xmath2 down to the @xmath28@xmath29 level .
what is needed are centimeter - level tracking data of a ( naturally or artificially ) drag - free satellite over time scales large enough to separate from newtonian contributions the secular motion of the eccentricity vector induced by a non - zero @xmath2 . among the class of zero - inclination ( equatorial ) orbits , geostationary ones
are nearly optimal .
along - track oscillations of a satellite with yearly period constitute another promising way of measuring @xmath2 around the @xmath29 level .
there is also a wide class of orbits for which preferred - frame effects on the eccentricity vector due to the orbital velocity of the earth around the sun are enhanced by small divisors .
these could be used to obtain the first bounds on @xmath2 independent of any hypothesis concerning the gravitationally preferred rest frame . in this paper
, we have presented an approximate analytical treatment of the dominant preferred - frame effects .
the value of this treatment is mainly indicative , as a help for selecting the most favorable orbits . in practice , we advise to resort to direct numerical integration of the equations of motion ( and , evidently , to a multiparameter fit to the experimental data ) . for the convenience of the interested reader , we end by giving the @xmath2 ( and for completeness @xmath3 )
contributions to the equations of motion .
contrary to the rest of the paper , we have in mind here global ( barycentric ) equations of motion ( written in a post - newtonian coordinate system appropriate to the description of the entire solar system ; see _ e.g. _ @xcite ) .
the relative acceleration of a satellite with respect to the earth has the form @xmath367 where the general relativistic geocentric acceleration ( including a relativistic treatment of multipolar and tidal effects ) will be found in full detail in ref .
@xcite , where @xmath368 denotes the acceleration induced by non - gravitational forces , and where ( to lowest order in the deviation from general relativity , and for @xmath369 , @xmath370 ) [ eq:58 ] @xmath371+{gm_\oplus\over r}\right){\bf n}_{\oplus\odot}\right\ } \ , \label{eq:58a}\\ { \bf a}_{\alpha_2 } = & & -\alpha_2\,{gm_\oplus\over 2r^2c^2}\left\ { ( { \bf n}\times{\bf w}_\oplus)^2\,{\bf n}+2({\bf n}\cdot{\bf w}_\oplus)\,({\bf n}\times{\bf w}_\oplus)\times{\bf n}\right\ } \nonumber\\ & & -\alpha_2\,{gm_\odot\over 2r_{\oplus\odot}^2c^2}\left\{{4\over 3}\,\left[{e^{\rm grav}_\oplus\over m_\oplus}\right]{\bf n}_{\oplus\odot}+{gm_\oplus\over r}\,({\bf
n}\times{\bf n}_{\oplus\odot})\times{\bf n}\right\ } \ . \label{eq:58b}\end{aligned}\ ] ] in these formulae , @xmath372 and @xmath373 denote the radial distance and the velocity of the satellite with respect to the earth , @xmath374 the unitary vector directed from the earth to the satellite , @xmath375 the unitary vector directed from the earth to the sun@xcite , and @xmath376 the absolute velocity of the earth with respect to the gravitationally preferred rest frame , @xmath32 being its orbital velocity around the sun . for conceptual clarity
we have indicated within square brackets the contributions of @xmath2 and @xmath3 due to the violation of the strong equivalence principle , _
i.e. _ the terms generated by expanding the @xmath11-dependent term coming from the lagrangian @xmath14 , eq .
( [ eq:2 ] ) .
the different terms of eqs .
( [ eq:58 ] ) are classified by order of decreasing magnitude .
[ the even smaller contributions due to the coupling of the satellite to the earth s intrinsic angular momentum are given in eq .
( 9.20 ) of ref .
@xcite ; they cause an additional secular precession of the satellite s orbit which is independent of the absolute velocity with respect to the preferred frame . ]
the contribution proportional to @xmath377 in eq .
( [ eq:58a ] ) is responsible for the along - track perturbations of the satellite studied in section v.b above , whereas the one proportional to @xmath378 , together with the contribution proportional to @xmath379 in the second line , is responsible for the perigee displacements studied in the rest of the paper .
note that the @xmath380 term has the same form as the one due to a violation of the strong equivalence principle [ with an opposite sign : @xmath381 , @xmath382 . actually , neither of these two terms is of much observational significance , as their contributions are much smaller than that due to the @xmath383 term . the force proportional to @xmath384 in the second line of eq .
( [ eq:58a ] ) has not been considered in the rest of the paper .
it arises in the connection between the locally inertial geocentric frame used in the body of the paper and the global barycentric one used here , and has the form of a coriolis force .
this coriolis force acts also on gyroscopes ( including the spinning earth ) and adds up to several other relativistic effects causing a universal precession @xmath385 of gyroscopes and satellite s orbits with respect to the barycentric frame ( see section 9.1 of @xcite and @xcite ) .
the @xmath2-dependent contribution to @xmath385 reads @xmath386 although the time dependence of @xmath387 , _ via _
@xmath388 , might in principle allow one to separate it from the other relativistic contributions to @xmath385 ( when discussing observables related to the global barycentric frame ) , its magnitude is too small to be of observational significance .
[ it yields a yearly oscillation of the satellite with respect to the barycentric frame which is smaller than @xmath389 . ] t. damour and g. esposito - farse , class .
quantum grav .
* 9 * , 2093 ( 1992 ) .
will and k. nordtvedt , astrophys
. j. * 177 * , 757 ( 1972 ) . c.m .
will , _ theory and experiment in gravitational physics _
( cambridge university press , cambridge , 1981 , revised 1993 ) ; int .
d * 1 * , 13 ( 1992 ) .
k. nordtvedt , astrophys .
j. * 320 * , 871 ( 1987 ) ; beware that the latter paper uses a nonstandard normalization for @xmath3 : @xmath390 . in the present paper , we shall limit our consideration to the four post - newtonian parameters @xmath15 , @xmath16 , @xmath2 , @xmath3 , with special attention to @xmath2 .
although extended versions of the parametrized post - newtonian ( ppn ) formalism have been defined , containing up to ten ppn parameters , the extra six parameters seem to be less well - rooted in plausible alternative field theories of gravitation .
k. nordtvedt and c.m .
will , astrophys .
j. * 177 * , 775 ( 1972 ) .
r.j . warburton and j.m .
goodkind , astrophys .
j. * 208 * , 881 ( 1976 ) .
hellings , in _ general relativity and gravitation , _ edited by b. bertotti , f. de felice and a. pascolini , ( d. reidel publishing company , dordrecht , 1984 ) , pp . 365385 . t. damour and g. esposito - farse , phys . rev .
d * 46 * , 4128 ( 1992 ) .
t. damour and k. nordtvedt , phys .
lett . * 70 * , 2217 ( 1992 ) ; and phys .
d , 15 october 1993 ( in press ) .
et al . _ ,
* 371 * , l1 ( 1991 ) .
actually , as we are interested in probing possible small deviations from general relativity ( @xmath391 , @xmath392 , ) it would ( iteratively ) suffice , even when dealing with orbital - period effects , to convert coordinate quantities into observable ones by interpreting them within a curved spacetime endowed with the _ general relativistic _
metric @xmath393 , without worrying about the @xmath2-dependence ( and @xmath394-dependence , ) of the coordinate / observable conversion .
we follow the standard astronomical conventions in denoting by @xmath395 the declination ( angle above the earth equator ; @xmath396 for the north pole ) and by @xmath397 the right ascension ( longitude around the earth polar axis counted from the vernal point )
. t. damour and g. schfer , phys . rev .
lett . * 66 * , 2549 ( 1991 ) .
d. brouwer and g.m .
clemence , _ methods of celestial mechanics _ ( london : academic press , new york , 1961 ) . actually , there is a subtlety in the application of the traditional method of variation of the elements because the disturbing function depends also on velocities .
one has the choice of working either with osculating elements " ( functions of positions and velocities ) or with contact " ones ( functions of positions and canonical momenta ) ; see v.a .
brumberg , _ essential relativistic celestial mechanics _ ( adam hilger , bristol , 1991 ) .
when considering _ first - order secular _ perturbations the two definitions give the same results , apart from a constant shift in the linear drift of @xmath51 coming from a different definition of an ( averaged ) newtonian orbital period .
within the present approach , the center of mass is defined using the standard newtonian formulas in the preferred frame . to simplify our rough analytical estimates
we assume that the @xmath398 improvement due to the number of individual measurements just compensates the difficulty of determining the center of a circle from a limited segment ( @xmath93 a quarter ) .
k. nordtvedt , phys .
d * 7 * , 2347 ( 1973 ) ; see eqs .
( 28),(29 ) .
smith and p.j .
dunn , geophys .
research lett .
* 7 * , 437 ( 1980 ) .
a. milani , a.m. nobili and p. farinella , _ non - gravitational perturbations and satellite geodesy _
( adam hilger , bristol , 1987 ) . c.m .
will , astrophys .
j. * 169 * , 141 ( 1971 ) .
t. damour , m. soffel and c. xu , _ general - relativistic celestial mechanics iv .
theory of satellite motion _
d , in press ; see also t. damour , m. soffel and c. xu , phys .
d * 47 * , 3124 ( 1993 ) for a discussion of the universal coriolis force ( in general relativity ) and the associated gravitational larmor theorem .
note that this convention is _ opposite _ to the notation @xmath399 used in ref . | as gravity is a long - range force , one might _ a priori _ expect the universe s global matter distribution to select a preferred rest frame for local gravitational physics . at the post - newtonian approximation , two parameters suffice to describe the phenomenology of preferred - frame effects .
one of them has already been very tightly constrained ( @xmath0 , 90% c.l . ) , but the present bound on the other one is much weaker ( @xmath1 , 90% c.l . ) .
it is pointed out that the observation of particular orbits of artificial earth satellites has the potential of improving the @xmath2 limits by a couple of orders of magnitude , thanks to the appearance of small divisors which enhance the corresponding preferred - frame effects .
there is a discrete set of inclinations which lead to arbitrarily small divisors , while , among zero - inclination ( equatorial ) orbits , geostationary ones are near optimal .
the main @xmath2-induced effects are : ( i ) a complex secular evolution of the eccentricity vector of the orbit , describable as the vectorial sum of several independent rotations ; and ( ii ) a yearly oscillation in the longitude of the satellite . |
a qubit is the elementary unit of quantum information , and can be realized with a variety of two - level systems , such as confined electron spins in a semiconductor nanostructure . for electron spin based qubits , universal quantum gates
can be realized using zeeman coupling and spin - spin exchange interaction .
@xcite the direct heisenberg exchange coupling between two electron spins is determined by the overlap of electron orbitals , and is thus a short - range nearest neighbor interaction . in order to implement quantum algorithms efficiently ,
quantum gate operations on remote qubits , i.e. , controllable long - range couplings , are needed .
various quantum data buses have been introduced to bridge this gap .
@xcite in this context , we have proposed to use the ground states of a strongly coupled spin chain as a quantum data bus , or a spin bus . @xcite we have shown that the parity of the spin bus can significantly alter the long - range effective couplings and entanglement between qubits that are coupled to the spin bus , @xcite and external fields can modify the form of the effective interaction between the attached qubits .
@xcite more recently , we have also shown that high - fidelity quantum state transfer can be achieved via such a spin bus .
@xcite an ideal quantum information processor has identical qubits , with precise control over couplings between qubits , and the qubits should be well isolated from their environment . however , in reality it is essentially impossible to create identical qubits based on artificial structures such as quantum dots and josephson junctions , and in a solid state environment there are normally several sources of qubit variance .
for example , the size of a quantum dot and the electron orbitals are largely determined by the gate structure and the applied gate voltages .
they can also be strongly influenced by factors such as the band structure of the host semiconductor and the random potential landscape due to modulation doping .
furthermore , the coulomb exchange coupling between spin qubits is determined by the exponentially small overlap of the electron orbitals , and controlled by the gate voltages .
small variations in gate voltages could thus cause large changes in the exchange coupling .
such deviations from the ideal value could lead to imperfect gate operations , and possible gate errors .
@xcite there are generally also very slow charge traps in a semiconductor heterostructure , where a trap can switch between two different charge distributions at a time scale much longer than the qubit operation time scales . while such a trap would probably be static during a quantum operation , it could modify the exchange coupling to a value that is different from the calibrated value .
similarly , via hyperfine interaction , environmental nuclear spins produce a local random magnetic field for a quantum dot confined electron spin qubit .
@xcite this field can be considered quasi - static in the context of a spin bus because its dynamics is much slower than the bus mediated gates . in short , in building a practical quantum information processor , deviations from calibrated values for various control parameters are inevitable .
it is thus necessary to know the engineering tolerance in the variation of parameters such as the spin - spin coupling and external magnetic fields . in this paper
we study how the capabilities of a spin bus are affected by static random variations in the exchange couplings between the bus node spins and the external magnetic fields experienced by the bus nodes .
specifically , we study how the bus spectrum , bus - qubit coupling , and bus - mediated qubit - qubit coupling are affected by these random but static variations of the system parameters .
the paper is organized as follows . in sec .
[ sec : review ] we discuss how the strongly coupled heisenberg chain can act as a spin bus when external qubits are weakly attached to it .
we derive the effective hamiltonians of the qubit - bus system up to second order . in sec .
[ sec : random ] , we show how the fluctuations in exchange couplings and external magnetic fields within the spin bus could affect the fidelity of the bus . finally , the summary and discussion are given in sec .
[ sec : summary ] . in the appendices
we discuss how to obtain the effective hamiltonians using a projection method , and present more detailed results on the bus spectrum .
in this section , we discuss how a strongly coupled uniform antiferromagnetic heisenberg spin chain can be used as a quantum data bus , or spin bus , which coherently connects remote qubits .
@xcite in appendix [ app : eff_hamil ] , using a many - body perturbation method based on the projection operator , we derive the effective hamiltonians for the qubit - qubit and qubit - bus couplings to first and second order .
we also calculate numerically relevant energy gaps , local magnetic moments , and the effective couplings for finite spin chains . and @xmath0 ( red circles )
are weakly coupled to an ( a ) odd - size or ( b ) even - size spin bus ( blue circles ) .
the right - hand panels show the corresponding effective hamiltonians in the low energy limit .
an odd - size spin bus in its ground state acts as an effective spin-@xmath1 particle denoted by @xmath2 .
it is coupled to the external qubits with strength @xmath3 at first order in the perturbation theory , and induces an rkky - like coupling @xmath4 between the external qubits at second order .
the even - size spin bus only mediates an rkky - like coupling between the external qubits at second order.,scaledwidth=45.0% ] the system we consider has two spin qubits @xmath5 and @xmath0 weakly attached to an open spin chain @xmath2 , as illustrated in fig .
the hamiltonian of the total system @xcite is @xmath6 in the ideal case with uniform exchange couplings and a uniform external magnetic field ( along the @xmath7 direction , @xmath8 ) , the heisenberg hamiltonian @xmath9 of the spin-@xmath1 chain is written as @xmath10 where @xmath11 is the spin operator of the @xmath12-th node of the chain , @xmath13 is the total number of spins in the chain , and @xmath14 indicates a uniform antiferromagnetic coupling between any two nearest neighbor spins at sites @xmath12 and @xmath15 . in this work , we will assume that a small or vanishing external magnetic field @xmath16 is applied to the spin chain ( the operation of a spin chain under a finite external magnetic field is discussed in refs . and .
hereafter , the spin chain @xmath2 will be referred to as a spin bus because it acts as a quantum data bus .
the hamiltonian @xmath17 describes antiferromagnetic couplings of qubits @xmath5 and @xmath0 to the @xmath12-th spin and the @xmath18-th spin of the chain , respectively @xmath19 we assume that the bus - qubit couplings @xmath20 with @xmath21 are small enough that the spin bus remains in its ground state manifold at all times . in this limit
, the total hamiltonian @xmath22 can be split into the unperturbed hamiltonian @xmath23 and the perturbation @xmath24 .
the perturbation condition is @xmath25 where @xmath26 is the zero - field gap @xcite above the ground state manifold ( for an odd - size bus ) or ground state ( for an even - size bus ) .
@xmath27 is thus used as a perturbation parameter .
the qubit - bus coupling @xmath20 can be turned on and off ( gradually ) , unlike the static intra - bus coupling @xmath28 . in general , external magnetic fields
@xmath29 and @xmath30 may be applied to qubits @xmath5 and @xmath0 to implement single - qubit operations on them , so that the hamiltonian @xmath31 can be written as @xmath32 there is no direct exchange coupling between qubits @xmath5 and @xmath0 , as they are nominally well separated . as shown later , the spin bus @xmath2 can mediate an effective coupling between them if they are both coupled to the bus .
since single - qubit operations are generally done separately from two- or multi - qubit operations , we set @xmath33 throughout this paper and will focus on qubit - bus and two - qubit couplings .
note that we will set @xmath34 and @xmath35 below for convenience . in the perturbative limit described by eq .
( [ perturbation_condition ] ) , we can perform a canonical transformation of the full hamiltonian in eq .
( [ total_hamil ] ) to obtain an effective hamiltonian where the spin bus is in its ground state manifold .
details of the transformation are provided in appendix a , as well as refs . and .
the actual form of the effective interaction depends on the parity of the bus .
an odd - size spin bus has a doubly degenerate ground manifold , and acts as an effective spin-1/2 particle . at first order in the perturbation theory , the spin-1/2 bus couples directly to the external qubits . at second order
, the spin bus mediates an effective rkky - like coupling between the qubits .
the resulting effective hamiltonian is given as follows : @xcite @xmath36 where @xmath37 is a spin operator representing the ground doublet states of the spin bus .
the effective coupling between qubit @xmath38 and the spin - bus @xmath2 is given to first order in the perturbation parameter by [ eq : coupling_j1 ] @xmath39 this is a product of the bare coupling @xmath20 between the @xmath12-th spin of the spin bus and the external qubit @xmath38 and the expectation value @xmath40 of @xmath41 at site @xmath12 in the ground state of the spin bus . notice that although @xmath40 is dimensionless , it can be considered as the local magnetic moment at site @xmath12 when multiplied by @xmath42 .
the rkky - like second - order coupling @xmath43 is given by @xcite @xmath44 here @xmath45 and @xmath46 are the eigenenergies and eigenstates of @xmath9 of an isolated spin bus , and @xmath47 with @xmath48 stand for pauli operators of the @xmath12-th spin of the spin chain .
the prime symbol on the summation indicates the exclusion of the ground states .
at zero external field , the ground state @xmath49 in eq .
( [ eq : j_rkky ] ) can be either @xmath50 or @xmath51 , or any linear combination between them .
the choice does not change the value of @xmath4 . at a finite magnetic field ,
however , if the ground state of the spin bus is degenerate , the two states are generally not spin - flipped image of each other . in this case , eq . ( [ eq : j_rkky ] ) has to be modified .
the effective hamiltonian ( [ eff_hamil_odd ] ) shows that the odd - size bus at zero or low field acts as an effective spin-@xmath1 particle that is coupled to the external qubits @xmath5 and @xmath0 , as illustrated in fig .
although in general @xmath52 , the second order term plays an essential role in long - time evolutions , such as in quantum state transfer .
@xcite thus our calculations in the rest of this paper are mostly concerned with these two coupling strengths .
furthermore , we focus on their normalized form @xmath53 and @xmath54 , which depend only on the size @xmath13 of the spin bus and the external magnetic field . for an even - size bus ,
the sub - hilbert space of interest is spanned by the non - degenerate ground state of the bus and the four eigenstates of the two qubits ( again we focus on the low - field limit ) . within this space
the bus does not have any dynamics as it is represented by a single ground state .
as for the two qubits , there is no first order effective coupling between them here , in contrast to the case of an odd - size bus . the second - order qubit coupling term is obtained in the same way as for an odd - size bus .
the effective hamiltonian to second order in the perturbation is given by @xcite @xmath55 where the rkky - like coupling @xmath4 has the same form as eq .
( [ eq : j_rkky ] ) . in this case
, the prime indicates that the non - degenerate ground state is excluded from the summation . based on the effective hamiltonians and the corresponding parameters
, we can make some qualitative observations on where the bus - qubit system might be susceptible to randomness and fluctuations . in the case of an odd - size spin bus with attached qubits ,
the key features that determine the operation of the bus include the zeeman splitting of the ground doublet of the spin bus , and the energy gap separating the ground doublet and the excited states , @xmath56 .
the former depends on the magnetic environment for the bus , while the latter depends on the interaction strength between the bus nodes .
both the qubit - bus couplings @xmath57 and the effective qubit - qubit couplings @xmath4 depend on the local exchange couplings and the local magnetic moments of the bus in its ground state manifold , which is a function of both magnetic environment and the intra - bus exchange couplings . in the case of an even - bus with attached qubits , @xmath4 has similar dependence on system environment as in the odd - size bus case , and is thus susceptible to variations in both the local magnetic fields and exchange couplings .
in sec . [ sec : review ] , we have shown how a heisenberg spin chain with uniform exchange coupling @xmath28 acts as a spin bus .
now we address the main question of the present paper , on how static randomness in exchange couplings and external magnetic fields can affect the fitness of the spin chain as a quantum data bus .
more specifically , we investigate how such randomness influences the two energy gaps , @xmath58 and @xmath56 , and the effective qubit - bus and qubit - qubit couplings @xmath57 and @xmath4 . in order to take into account the effects of randomness in exchange couplings and applied magnetic fields , the hamiltonian of the chain ( [ chain_hamil ] ) is generalized to @xmath59 where @xmath60 is the antiferromagnetic coupling between two neighboring spins at sites @xmath12 and @xmath15 , and @xmath61 is the local magnetic field at the @xmath12th site of the spin bus . note that in spite of the random @xmath62 and @xmath61 , it can be easily shown that hamiltonian ( [ hamil_random ] ) still commutes with the @xmath7 component of the total spin @xmath63 , i.e. , @xmath64=0 $ ] .
hamiltonian ( [ hamil_random ] ) may be considered as a finite quantum spin glass model .
@xcite there are several spin glass models depending on the types of couplings ( ising or heisenberg , and short range or long range ) and the distributions of @xmath62 .
for example , the sherrington - kirkpatrick model @xcite has the couplings between arbitrary pairs , sampled from the normal distribution with zero mean , while the edwards - anderson model @xcite has only the nearest neighbor couplings . in the context of quantum information processing
, we can reasonably assume that the exchange couplings and the applied magnetic fields are both near their target values , @xmath28 and @xmath16 .
the random exchange couplings @xmath62 and applied magnetic fields @xmath61 are then [ random_j&b ] @xmath65 in the numerical analysis described below , we choose @xmath66 and @xmath67 that are randomly sampled from normal distributions with standard deviations @xmath68 and @xmath69 , respectively . in experimental systems
, we would expect such variations to be small , assuming reasonable calibration efforts . in the following studies ,
we analyze these two types of random variations separately , keeping one of the variables uniform . in this subsection , we investigate how random variations in the inter - node exchange couplings @xmath62 , given by eq .
( [ random_j&b]a ) , affect the ability of an odd - size chain to function as a spin bus .
such variations could result from calibration errors , slow but random hopping of charge traps near the spin bus nodes , and whatever other factors that are not accounted for during the calibration process . here
the external magnetic field @xmath16 on the chain is set to be zero or small , so that the system remains in the isotropic regime .
the external magnetic field , if any , is taken to be uniform , so that @xmath70 .
the variations @xmath66 are small compared to @xmath28 , so that they may be treated as a perturbation : [ eq : hamil_random_j ] @xmath71 where @xmath72 is the unperturbed hamiltonian ( [ chain_hamil ] ) , and the perturbation @xmath73 is given by @xmath74 hereafter the superscript @xmath75 is used to denote the case of uniform exchange coupling or uniform magnetic fields in the bus . and
@xmath76 , and ( b ) two energy gaps , @xmath77 and @xmath78 are plotted as a function of sample number @xmath79 of odd - size chains with size @xmath80 with random exchange couplings @xmath62 with the standard deviation @xmath81 at @xmath82 . in ( c ) and
( d ) , two energy gaps @xmath58 and @xmath56 are plotted as a function of @xmath16 for the samples @xmath83 . ] here we consider an ensemble of @xmath79 odd - size chains .
each of the samples in this ensemble has the same size @xmath13 but different @xmath62 sampled from the normal distribution with average @xmath28 and standard deviation @xmath68 .
[ fig : odd_rj_gap](a ) shows the fluctuations in energy levels as a function of sample number , at a low magnetic field of @xmath82 .
the two lowest energy levels @xmath84 and @xmath85 fluctuate in sync , so that the gap @xmath58 is free from the randomness of @xmath62 [ shown in fig .
[ fig : odd_rj_gap](b ) ] . the energies @xmath86 and @xmath76 of @xmath87 and @xmath88 are also in sync , as shown in fig .
[ fig : odd_rj_gap ] ( a )
. however , the gap @xmath56 , which is a measure of the isolation of the ground doublet from the excited states , does fluctuate [ as shown in fig .
[ fig : odd_rj_gap](b ) ] , because @xmath85 and @xmath86 have different dependence on the exchange coupling . in other words , while the ground state splitting @xmath58 of this odd - size bus is robust against the randomness in exchange coupling , the ground - excited - state gap @xmath56 is affected by the randomness . in figs .
[ fig : odd_rj_gap ] ( c ) and ( d ) the two gaps , @xmath58 and @xmath56 , are plotted as a function of the uniform magnetic field @xmath16 applied on the bus . at low fields ,
the ground state gap @xmath58 increases linearly and without broadening as the magnetic field increases , until @xmath89 .
it starts to be influenced by the exchange randomness above @xmath89 , which corresponds to the crossing between levels @xmath90 and @xmath91 , as shown in fig .
[ fig : odd_gap ] in appendix [ app : odd_bus ] . beyond this crossing point , states @xmath92 and @xmath93 are not the time - reversal of each other anymore due to the level crossings with higher excited states .
recall that for a spin chain to act as a spin bus , we need the ground state doublet to be well separated from the excited states , or @xmath94 .
panels ( c ) and ( d ) of fig .
[ fig : odd_rj_gap ] indicate that this condition is satisfied when the bus ground doublet is energetically separated from excited states ( @xmath95 ) , and acts as an effective spin-1/2 system with a constant magnetic moment .
although @xmath64 = 0 $ ] dictates that the dimensionless total magnetic moment @xmath96 of the ground state is still a good quantum number despite random exchange couplings , the local magnetic moment @xmath40 does fluctuate around @xmath97 , as shown in fig [ fig : local_moment ] .
consequently , the first - order effective coupling @xmath98 between the qubit and the bus , given by eq .
( [ eq : coupling_j1 ] ) , is affected by the randomness in the bus exchange coupling , and has to be calibrated individually . are plotted ( red dots ) for all lattice sites on an odd - size spin bus with @xmath99 for an ensemble of @xmath100 random samples , with @xmath101 and @xmath102 .
the height of the blue bars represents the local magnetic moment @xmath97 for uniform exchange couplings ( @xmath103 ) .
the spread of the red dots at each bus node illustrates the variations in local magnetic moments due to the random exchange couplings , with larger variations apparent at the end sites . ]
the effective qubit - qubit coupling @xmath4 is also affected by the randomness of the intra - bus exchange couplings @xmath62 . as indicated in eq .
( [ eq : j_rkky ] ) , @xmath4 is determined by the full bus spectrum , including both the energy levels and the excited states .
figures [ fig : odd_rj_gap ] and [ fig : local_moment ] show that the random exchange couplings @xmath62 in general affect the energy gaps from the ground state , @xmath104 , as well as the bus eigenstates @xmath46 .
thus we expect that @xmath4 should be sensitive to the randomness in @xmath62 .
figure [ fig : odd_random_j ] shows how the ensemble averages of the gap , @xmath105 , the local magnetic moment , @xmath106 ( which is the normalized first - order qubit - bus coupling @xmath107 ) , and the normalized second - order effective coupling , @xmath108 depend on the fluctuations of the exchange coupling , represented by the standard deviation @xmath68 , over a 5-node bus . the blue filled circles in fig . [
fig : odd_random_j ] show how the fluctuations in these quantities depend on the randomness in the exchange couplings .
for example , when @xmath109 , @xmath110 , which indicates that the effective qubit - bus coupling and the effective qubit - qubit coupling have similar sensitivities to the random variations in the intra - bus exchange couplings .
furthermore , both the local magnetic moment ( thus the qubit - bus coupling ) and the effective qubit coupling are linear functions of @xmath68 , with their slopes depending on the size of the bus .
these slopes are indicators of sensitivity of @xmath111 and @xmath43 to the exchange variations , and can be used in evaluating decoherence in such a spin bus architecture .
for example , background charge fluctuations can affect exchange couplings between neighboring nodes of a spin bus . as a result ,
the effective qubit - qubit exchange coupling becomes a time - dependent random variable , which leads to two - qubit dephasing .
@xcite the relevant correlation function that determines the dephasing is @xmath112 , and is given approximately by @xmath113 ^ 2 { \langle j_i(t ) j_i(0)\rangle}$ ] @xcite .
the latter correlation function , @xmath114 , represents fluctuations in the individual inter - node exchange couplings along the bus , whose dynamics is determined by the environmental charge noise . , the ensemble averages of ( a ) the gap @xmath105 in units of @xmath28 , ( b ) the dimensionless local magnetic moment @xmath106 at the end of the chain , and ( c ) the normalized second - order effective coupling @xmath115 are plotted ( open red circle ) as a function of the standard deviation @xmath68 of @xmath66 .
the blue filled circles in each panel indicate the standard deviations of each data point , which is obtained by averaging over 2000 random configurations . here
@xmath116 . ] in summary , even when intra - bus exchange couplings of an odd - size spin chain have random but static variations , the chain can still act as a spin bus , with a ground state doublet that is well separated from the excited states , and acts as an effective spin-1/2 system with a constant magnetic moment . however , the effective qubit - bus couplings and the mediated qubit - qubit couplings are affected by the randomness in exchange , with their fluctuations linearly proportional to the randomness in exchange . calibration
would thus be needed for accurate qubit operations .
the results here also have implications for spin - bus related decoherence .
in essence , the strong exchange couplings allow a spin bus to process quantum information across a large distance , but also make the qubit - bus system susceptible to charge noise via both @xmath117 and @xmath4 .
spin qubits are generally susceptible to magnetic noise , and the spin bus is no exception .
here we examine how random but static external magnetic fields affect the properties of an odd - size spin bus .
our results should also be a useful indicator of the sensitivity of a spin bus to temporal magnetic noise , as we will discuss later in the section .
for this calculation we assume that the exchange couplings @xmath62 are uniform , and focus on the magnetic randomness . the local magnetic fields , eq .
( [ random_b ] ) , are @xmath118 where the random field @xmath67 is sampled from a normal distribution with standard deviation @xmath69 . as in the previous subsection , we consider an ensemble of @xmath79 spin chains , so that the ensemble average of local magnetic field @xmath61 is @xmath119 and @xmath120 in the limit of large @xmath79 .
such a random distribution of local magnetic field could originate from quasi - static nuclear hyperfine fields , or local paramagnetic centers in a semiconductor . in units of @xmath28 between the two lowest states of a single spin as a function of the external magnetic field @xmath16 .
the height and the width of the rectangle , touching the bottom of the curve , indicates @xmath121 and @xmath122 at @xmath102 , respectively . ] as a benchmark , we first recall how an odd - size spin chain behaves in a constant uniform magnetic field . when @xmath116 , the odd - size chain has two doubly degenerate ground states with the total magnetic moment @xmath123 .
a finite @xmath16 splits these two states like a single spin-@xmath1 particle .
there are two important features that determine the behavior of the ground doublet in a magnetic field : the energy splitting @xmath58 , and the ground state spin orientation . in a uniform field
the latter depends only on the @xmath124-factor of the material , while in a random field it also depends on the local field configuration .
we first examine how the random external magnetic field affects @xmath58 , the splitting of the ground state doublet of the odd - size bus .
this dependence is a good indicator of whether the ground state doublet of an odd - size spin bus is a robust effective spin-1/2 .
note that the commutation relation @xmath64 = 0 $ ] holds even in a random magnetic field , so the two lowest states have @xmath123 ( recall that the inter - node coupling is anti - ferromagnetic ) .
figure [ fig : odd_randomb_gap ] shows that the randomness in the magnetic field induces a finite average gap at zero field .
this gap opens at @xmath116 even though @xmath120 .
this non - vanishing average gap between the two lowest states is a consequence of the statistical behavior of the random field , and can be understood using a model of @xmath79 single spins in a gaussian ensemble of random external magnetic fields @xmath125 with @xmath116 .
the hamiltonian of a single spin is @xmath126 the energy splitting of each spin is given by @xmath127 , which is _ always _ positive .
it is thus not a surprise that the ensemble average of the gaps , @xmath128 , is nonzero , even though @xmath129the ground state changes according to the field configuration . to make this argument more rigorous ,
recall that for a normal distribution , the odd central absolute moments of a random variable @xmath130 with a mean of @xmath131 are given by @xmath132 where @xmath133 denotes the double factorial . applying eq .
( [ odd_expect ] ) to @xmath58 , we get @xmath134 where @xmath69 is the standard deviation of @xmath125 .
while in our case the magnetic moment of the ground state is distributed throughout the whole spin chain , its splitting is still _ mainly _ due to the zeeman splitting of a single bohr magneton , so that the single - spin argument provided here is still applicable .
when @xmath135 , the average gap can be qualitatively expressed as @xmath136 , which is a hyperbola that saturates at @xmath116 to @xmath137 and approaches @xmath138 at large @xmath16 .
the difference in the proportionality constant originates from the difference between @xmath139 and @xmath140 . of the ground state @xmath49 , ( b ) local magnetic moment @xmath141 for the ground state @xmath49 ( orange solid circle ) and for the first excited state @xmath142 ( green open circle ) , and ( c ) the fidelity @xmath143 of the ground state in the presence of random magnetic fields with respect to that in the uniform field as a function of the sample number . here
the size of spin buses is @xmath99 , @xmath144 , and @xmath145 .
] random realizations of odd - size chains with @xmath146 , as a function of @xmath16 and @xmath69 . ]
now we address the spin orientation of the odd - size spin bus in the ground state under random magnetic fields .
the magnetic moment of a single spin ( with @xmath147 ) in the ground state is anti - parallel to the external field . as in the single spin case ,
the total magnetic moment of the odd - size spin bus in the ground state is always anti - parallel to a _ small uniform _ external magnetic field . as shown in fig .
[ fig : local_moment ] , the local magnetic moments align antiferromagnetically ( alternation between anti - parallel and parallel alignments ) . even in a random external magnetic field in the @xmath7-direction ,
the two lowest states of an odd - size spin bus are characterized by @xmath148 or @xmath149 , so one might guess the zeeman energy would determine the spin orientation .
however , this is not the case .
figure [ fig : rand_b_moment ] shows the total magnetic moment of the ground state @xmath49 in an ensemble of random magnetic fields for the case when @xmath150 and @xmath151 . as illustrated in panels ( a ) and ( c ) , seven samples , # 10 , # 19 , # 36 , # 44 , # 57 , # 69 , and # 97 , among 100 samples have negative total magnetic moment @xmath152 , so that their ground state is @xmath153 instead of @xmath154 .
panel ( b ) of fig .
[ fig : rand_b_moment ] shows a strongly random local magnetic moment on the 9-node spin bus , with the seven flipped states having reversed local magnetic moments . the probability for either @xmath153 or @xmath154 to be the ground state depends on the ratio of @xmath16 and @xmath69 , as illustrated in fig .
[ fig : odd_rb_sigma ] . as expected ,
when @xmath116 , the probability is 50% .
a strong uniform field @xmath16 ( compared to @xmath69 ) suppresses the flipping fraction and stabilizes one of the states as the predominant ground state . a close inspection of the data presented in fig .
[ fig : rand_b_moment ] reveals that zeeman splitting of the local nodes of a spin bus does not tell the whole story of @xmath155 , so that the single - spin model has its limitations . by definition ,
the energy gap is @xmath156 here @xmath157 and @xmath158 represent the exchange and zeeman components of @xmath9 in eq .
( [ hamil_random ] ) , @xmath159 and @xmath160 denote the zeeman and exchange contributions to the energy of state @xmath161 , with @xmath162 , @xmath163 is the local magnetic moment at bus node @xmath12 , and we have taken @xmath164 for simplicity . when the external magnetic field is uniform , @xmath165 , the exchange contribution to the energy of @xmath49 and @xmath142 in eq .
( [ eq : energy_gap ] ) are the same , so that @xmath166 , and @xmath167 are equal in magnitude and opposite in sign to @xmath168 .
thus the gap between the two states is given completely by the zeeman splitting : @xmath169 . with @xmath147 , the net spin of
the ground state @xmath49 is parallel to the uniform magnetic field @xmath170 like a single spin , so that @xmath171 is the ground state when @xmath0 is positive . under a random magnetic field @xmath61 , but with @xmath172 , the exchange contribution to the gap in eq .
( [ eq : energy_gap ] ) is still negligible , and the ground state spin is parallel to @xmath16 .
however , if @xmath16 is of the order of @xmath69 or smaller , the ground state total spin orientation may be anti - parallel rather than parallel to @xmath16 .
for these configurations with the `` flipped '' ground state , the reason for the flipping is varied .
consider the samples ( from the ensemble of 9-node buses presented in fig .
[ fig : rand_b_moment ] ) shown in fig .
[ fig : n9rb_flip ] and table [ tab2 ] . here
sample # 0 refers to the spin bus in a uniform magnetic field .
samples # 6 , # 10 , # 19 and # 36 are for the bus in different random magnetic field configurations , and the later three samples have a flipped ground state , with @xmath173 .
figure [ fig : n9rb_flip ] shows that in a random magnetic field , the local magnetic moments for the two lowest states are generally not equal in magnitude , @xmath174 , so that the two states @xmath49 and @xmath142 are no - longer spin - flipped image of each other , even though @xmath175 .
the most dramatic examples are those samples with a flipped ground state , where the ground and first excited states are far from the classical anti - ferromagnetic spin configurations .
this also implies that in general the zeeman energy @xmath176 .
samples # 6 and # 10 demonstrate that even if all the magnetic fields @xmath61 are positive , the ground state could still be either @xmath148 or @xmath177 . while in most cases the energy gap between the two lowest states in eq .
( [ eq : energy_gap ] ) is due to the zeeman term , for samples # 10 and # 19 the zeeman energies @xmath178 and @xmath179 are almost equal , so that there is no zeeman gap . the energy gaps for these two samples are determined by the exchange contribution in eq .
( [ eq : energy_gap ] ) , and are one or two orders of magnitude smaller than the usual zeeman gap . for sample # 36 , the total energy gap is dominated by the zeeman contribution ( with significant contribution also coming from the exchange interaction ) , although the configuration of the random magnetic field is such that the ground state is flipped . in short , when @xmath180 , the local and total magnetic moments of the ground state are sensitively dependent on the distribution of the random magnetic field .
.(color online ) energy gap @xmath181 between the bus states @xmath49 and @xmath142 , its zeeman contribution @xmath182 , the exchange contribution @xmath183 , and the absolute ratio of these contributions for samples # 0 , # 6 , # 10 , # 19 , and # 36 in fig .
[ fig : rand_b_moment ] . here
the energy is measured in units of @xmath28 . [ cols="^,^,^,^,^",options="header " , ] our results so far indicate that random magnetic fields can seriously undermine the capabilities of a spin bus , by altering the local magnetic moments ( and thus the effective qubit - bus coupling @xmath111 ) and the bus ground state .
fortunately , in general @xmath69 is relatively small .
for example , if the random field is due to hyperfine interaction in gaas , @xmath184 mt for a 100 nm quantum dot , so that a @xmath185 mt should be more than enough to overcome the effect of the random field . furthermore , the above discussion is applicable to the regime where the magnetic energy scales are not much smaller than the exchange energy scales ( e.g. , for fig .
[ fig : rand_b_moment ] and table [ tab2 ] , @xmath186 ) .
if @xmath187 and @xmath69 , the structures of the ground and first excited states of the spin chain should be determined by the anti - ferromagnetic coupling and are less susceptible to the small magnetic field or its fluctuations . in this case the ground state spin orientation would be mostly determined by the zeeman contribution to the energies of the two states , and we would recover the simple single - spin physical picture . .
the total magnetic moment of the ground states of samples # 0 and # 6 is @xmath188 , i.e. , @xmath189 .
on the other hand , samples # 10 , # 19 , and # 36 have the spin - flipped ground states , @xmath173 , i.e. , @xmath190 . ] the fluctuations in the local magnetic moments of the spin bus due to the random external magnetic field lead directly to fluctuations in the effective qubit - bus couplings @xmath117 [ see eq .
( [ eq : coupling_j1 ] ) ] , though this fluctuation is suppressed if @xmath172 , as illustrated by the few data points for small @xmath191 in panel ( b ) of fig .
[ fig : odd_random_b ] .
in addition , under a random magnetic field , the effective coupling @xmath98 , given by eq .
( [ eq : coupling_j1 ] ) , becomes anisotropic .
this is in addition to the anisotropy induced by a finite @xmath16.@xcite in general , anisotropy occurs when @xmath192\,,\end{aligned}\ ] ] and @xmath193 the anisotropy appears in both @xmath111 and @xmath43 @xcite , as illustrated in panels ( b ) and ( c ) of fig .
[ fig : odd_random_b ] . based on the standard deviation data presented in panel ( b )
, we also observe that while the transverse component of the local magnetic moment is reasonably robust against the randomness in the magnetic field , the longitudinal component is not . on the other hand , panel ( c )
shows that both the longitudinal and transverse components of the effective qubit - qubit coupling @xmath43 have a linear dependence on the field randomness for small @xmath69 , changing to a different slope as @xmath69 becomes larger than @xmath16 .
both of these observations illustrate the fact that for an odd - size spin bus to function properly , magnetic field randomness in the system needs to be minimized . , the ensemble averages of ( a ) the gap @xmath194 in units of @xmath28 , ( b ) the local magnetic moment @xmath106 on one end of the chain , and ( c ) the normalized second - order effective coupling @xmath195 are plotted as a function of the standard deviation @xmath69 of @xmath67 . in ( b )
, the labels `` x '' ( symbol : red open circle ) and `` z '' ( symbol : orange open diamond ) stand for the @xmath196- and @xmath7- components of the first - order effective coupling , respectively . in ( c )
the labels `` xx '' ( symbol : red open circle ) and `` zz '' ( symbol : orange open diamond ) refer to the @xmath197- and @xmath198-component of the second - order effective coupling , respectively . here
@xmath199 . in panels
( b ) and ( c ) , the solid diamonds and solid circles represent standard deviations for the data labeled by the open diamonds and open circles , respectively . ] as explained in sec .
[ sec : review ] , the effective qubit - qubit coupling @xmath43 given by eq .
( [ eq : j_rkky ] ) is well defined when the degenerate or nearly degenerate ground states of the spin bus are spin - flipped states of each other .
this is the case for an odd - size spin bus near zero external magnetic field . at certain finite external fields , the ground states of the spin bus
would be close to be degenerate again .
however , in those regimes eq .
( [ eq : j_rkky ] ) is generally not applicable and should be modified .
@xcite in summary , the effects of the random magnetic field in the @xmath7-direction on an odd - size bus are as follows .
first , the degeneracy in the ground states of an odd - size bus is lifted .
second , although the two lowest states have either @xmath148 or @xmath149 , the spin orientation of the ground state is not solely determined by the zeeman energy .
third , the random magnetic fields make the first - order and the second - order effective couplings anisotropic . in this subsection
, we investigate how an even - size spin bus is affected by random exchange couplings @xmath62 .
recall that the effective hamiltonian for an even - size bus coupled to two qubits in zero magnetic field takes the form of @xmath200 , where @xmath4 is given by eq .
( [ eq : j_rkky ] ) .
the various terms in this equation are the bare qubit - bus couplings @xmath201 and @xmath202 , the bus energy gaps @xmath203 , and the transition matrix elements @xmath204 . here
we examine the effects of random intra - bus exchange couplings on these different terms , with a particular focus on the gap @xmath58 between the ground and the first excited state , which is an indicator of how well the nondegenerate ground state is isolated from the excited states , and figures prominently in the expression of @xmath43 .
an even - size bus can be thought of as an odd - size bus plus an extra spin-1/2 node .
thus the lowest four states of an even - size bus is a singlet and a triplet .
a uniform magnetic field would split the triplet , but would not affect the singlet ground state . with random exchange couplings , similar to the case of an odd - size bus , we still have @xmath205 = 0 $ ] , so that the triplet splitting is given by zeeman splitting .
the random exchange couplings do cause the energy levels and the eigenstates to vary in general .
for example , fig .
[ fig : even_rj_gap ] plots the two energy gaps , @xmath58 and @xmath56 , as a function of @xmath16 for 100 samples of even - size chains with random exchange couplings .
while the ground state remains nondegenerate , the gap @xmath58 is now distributed between 0.3 and @xmath206 at @xmath102 . around @xmath102 ,
the gap @xmath56 and the next gap @xmath207 ( not plotted ) are robust against the random exchange couplings @xmath62 , reflecting the fact that these two gaps correspond to the zeeman splittings of the triplet bus states .
this is similar to the @xmath58 gap of an odd - size bus near zero field , which corresponds to the zeeman splitting of the spin-1/2 doublet ground state , as shown in fig .
[ fig : odd_rj_gap ] . even - size chains with random exchange couplings , two lowest energy gaps , ( a ) @xmath208 and ( b ) @xmath209 , are plotted as functions of @xmath16 .
the size of each chain is @xmath210 and the standard deviation @xmath211 of @xmath66 is 0.025 . ]
the most important effect of the randomness in the exchange couplings for bus operations is on the low lying energy gaps such as @xmath58 . as shown in panel ( b ) of fig .
[ fig : even_rj_coupling ] , the ground state itself is quite robust against the random exchange coupling in terms of state fidelity , and has zero local magnetic moments as well as zero total magnetic moment .
this is in contrast with the effect of random magnetic fields on the even - size bus as shown in subsection [ subsec : even_randomb ] , where the local magnetic moments become non - zero . on the other hand , while the average value of the @xmath58 gap is only weakly dependent on @xmath68 , the fluctuations in @xmath58 depend linearly on @xmath68 , as shown in panel ( a ) of fig .
[ fig : even_rj_coupling ] .
consequently , the fluctuations of the bus - mediated qubit - qubit coupling also has a linear dependence on the standard deviation @xmath212 of @xmath66 , as shown in panel ( c ) of fig .
[ fig : even_rj_coupling ] .
the slope @xmath213 is quite large here , reflecting a sensitive dependence of the singlet - triplet splitting of the spin bus on the intra - bus exchange couplings .
while calibration should be able to largely suppress the effects of any static randomness of the exchange coupling , the sensitivity to randomness in the local exchange , as indicated by the large @xmath213 , dictates that effects of environmental charge noise on the inter - node exchange coupling @xmath62 need to be minimized .
, ( a ) the ensemble averages of the ground energy gap @xmath58 , ( b ) the fidelity @xmath143 with respect to the uniform ground state , and ( c ) the normalized effective qubit - qubit coupling @xmath214 are plotted as functions of the fluctuation @xmath68 of the random exchange coupling ( all data are represented by red open circles ) . here
@xmath116 and the sample size is @xmath215 .
the blue solid circles represent the standard deviation of @xmath58 , @xmath143 , and @xmath216 , respectively . ] in this subsection , we study how an even - size bus is affected by an external magnetic field that has random local variations . in a uniform but small external magnetic field ,
the ground state of the even - size bus has zero local magnetic moments as well as zero total magnetic moment ( @xmath217 ) .
recall that the commutation relation @xmath64=0 $ ] is still valid even when the bus is subject to random magnetic fields @xmath61 in the @xmath7-direction .
if the random magnetic fields are weak , the even - size bus remains in the ground state with @xmath217 , i.e. , zero total magnetic moment .
however , the local magnetic moments @xmath40 become non - zero in contrast with the case of a uniform magnetic field .
furthermore , the bus excited states generally do have net magnetic moments , so that they respond to both local and global magnetic fields in terms of their energies and their state composition . .
( a ) the average ground energy gap @xmath58 in units of @xmath28 , ( b ) the average fidelity @xmath143 of the ground states with respect to that without random @xmath218 , and ( c ) the average of the normalized second - order effective coupling @xmath216 are plotted ( in open red circles or open orange diamonds ) as a function of the standard deviation of the random magnetic field @xmath69 .
the standard deviations of the averages are given by the blue solid circles or blue solid diamonds , respectively .
here we take @xmath116 . ] in fig .
[ fig : even_rb ] we plot the effect of the locally random magnetic field on the gap @xmath58 , the ground state robustness ( in terms of the fidelity @xmath143 with respect to the uniform field ground state ) , and the bus - mediated qubit - qubit interaction @xmath219 .
the average of gap @xmath58 decreases with @xmath69 because for any particular random field configuration , one of the polarized triplets has a lower energy compared to the zero field case and becomes the first excited state .
the linear increase in the standard deviation of @xmath58 is simply a reflection of the linear nature of the zeeman splitting . in panel ( b ) , the decrease in the fidelity shows that the local magnetic moments may fluctuate although their sum , i.e. , the total magnetic moment is still zero .
interestingly , in panel ( c ) , the normalized effective coupling @xmath220 and its standard deviation both have a super - linear dependence on @xmath68 when it is small , making the coupling a robust quantity against field randomness . the qualitative reason for this robustness
is that the ground state is not coupled to the two polarized triplet states of the bus by the random magnetic field along @xmath7 direction at the lowest order , while the gap between the unpolarized triplet state and the ground state only depends on the field randomness quadratically . in this calculation
we did not consider a random transverse field since our focus is on static disorder .
for example , for magnetic disorder caused by random nuclear polarizations , the transverse polarizations would precess around the external field , so that their effect would tend to be suppressed . in ref .
, we have shown that a constant external magnetic field makes the second - order effective interaction mediated by an even - size bus anisotropic . here
we find that local random variations in the magnetic field , @xmath67 , also induce anisotropy , even if @xmath102 and @xmath221 , as shown in panel ( c ) of fig .
[ fig : even_rb ] .
this anisotropy is weaker compared to the case of an odd bus in a finite @xmath16 as shown in fig .
[ fig : odd_random_b ] , because the average of the local field here vanishes ( @xmath116 ) , so that the spatial isotropy is not broken as completely as in the case of fig .
[ fig : odd_random_b ] . note that we would normally expect an even - size bus to be operated at zero external magnetic field @xmath102 , unless the anisotropy of the effective coupling is desired .
we have performed a comprehensive study of the effects of local randomness in the exchange couplings and the external magnetic field on the capabilities of a strongly coupled heisenberg spin bus .
we find that the random exchange couplings preserve the isotropic symmetry ( in the qubit - bus and qubit - qubit couplings ) of the bare heisenberg coupling .
this symmetry also makes the ground zeeman energy gap of the odd - size bus robust against small fluctuations in the exchange couplings .
however , randomness in the exchange couplings does cause the eigenenergies and the eigenstates to vary , which in turn leads to randomness in the magnitudes of the effective couplings ( both qubit - bus and qubit - qubit )
. an external magnetic field , whether uniform or random , does break the isotropy of the heisenberg spin bus , and leads to anisotropy in the qubit - bus and qubit - qubit effective couplings . a locally random magnetic field
also lifts the ground state degeneracy of an odd - size bus , even when the average applied field vanishes .
the local randomness also gives rise to the effect that the total magnetic moment of the odd size bus in the ground state may be _ antiparallel to the direction of the applied magnetic fields _ , when a single spin in the ground state would have been parallel to the magnetic field .
even - size buses are somewhat more robust against local random magnetic fields , since their ground state is non - magnetic .
we have performed ensemble calculations for the coupled qubit - bus systems we have considered , where the standard deviation of an ensemble averaged quantity ( such as the qubit - bus and qubit - qubit effective couplings ) represents the sensitivity of this quantity to the particular parameter randomness .
thus our results have clear implications not only for situations where static parameter randomness is present , but also for dynamical noise in the exchange coupling or the external field . to give context to this paper , in the appendices we provide a comprehensive overview of even and odd spin chains as quantum data buses .
in particular , we derive the first- and second - order effective couplings using the projection operator method .
we explore the low - energy spectra of buses coupled to zero , one or two qubits , from which we derive the first- and second - order effective hamiltonians of the qubit - bus system .
we also prove that random exchange couplings do not lift the ground state degeneracy of an odd - size bus .
finally , we present a study of the scaling properties of the bus , for up to 20 nodes .
this work was supported by the darpa quest program through afosr and nsa / lps through aro .
in this appendix we derive the effective low - energy hamiltonian of the full hamiltonian ( [ total_hamil ] ) . the weak qubit - bus couplings @xmath222 are used as perturbation parameters .
the total hamiltonian @xmath223 can be rewritten as an unperturbed hamiltonian @xmath23 and a perturbation @xmath24 .
since we are interested in the low - energy limit , we define a projection operator @xmath224 onto the subspace @xmath225 spanned by the tensor products @xmath226 of the ground state(s ) of the free bus hamiltonian @xmath9 and the eigenstates of the free qubit hamiltonian @xmath227 @xmath228 here the zeeman energy of @xmath227 is assumed to be small compared to the gap of @xmath9 , so that the energy of the subspace @xmath225 is equal or nearly equal to the ground energy @xmath84 of @xmath9 . in this limit
, we can set @xmath33 .
the finite - field induced anisotropy is addressed elsewhere .
@xcite the effective hamiltonian acting on the subspace @xmath225 is then given by @xcite @xmath229 where @xmath230 projects onto the sub - hilbert - space orthogonal to @xmath225 .
the effective hamiltonian to second order in @xmath231 is given by @xmath232 the derivation of the explicit form of the effective hamiltonian ( [ eff_hamil ] ) requires detailed information on @xmath224 , which consists of the structure and spectrum of the ground manifold of @xmath9 . the heisenberg spin chain hamiltonian @xmath9 , given by eq .
( [ chain_hamil ] ) , is exactly _ albeit only partially _ solvable with the bethe ansatz .
@xcite the @xmath7 component @xmath233 of the total spin @xmath234 commutes with the hamiltonian ( [ chain_hamil ] ) , @xmath235 = 0 $ ] , so that the energy eigenstates can be labeled by @xmath236 with the energy level @xmath237 and the magnetic quantum number @xmath238 , i.e. , the eigenvalues of @xmath233
. however , the general analytic expressions of the eigenstates @xmath236 are not available .
traditionally , for bulk systems , the periodic boundary condition and an even number @xmath13 ( @xmath239 in thermodynamic limit ) are assumed . for finite size chains , however , _
the eigenstates are dependent on both the boundary condition and the even - odd parity of size @xmath13 _ , as shown in figs .
[ fig : odd_eng ] and [ fig : even_eng ] .
thus , the effective hamiltonians for the bus - qubit system are different depending on the parity of the bus.@xcite here we give a more detailed description of the derivation of the effective hamiltonians . note
that ref . demonstrated experimentally the even - odd parity effect of a spin chain , by assembling chains of 1 to 10 mn atoms on a metallic surface and measuring the parity dependent tunneling currents .
an odd - size antiferromagnetic chain has an odd number of spins , so that the ground state should have one uncompensated spin .
for example , for @xmath240 , the classical antiferromagnetic spin configurations are `` up - down - up '' or `` down - up - down '' .
the exact degenerate quantum mechanical ground states of the odd - size chain with @xmath240 around @xmath102 are given by [ 3state ] @xmath241 where @xmath49 and @xmath142 on the right hand side represent the spin up and down states of a single spin , respectively .
one can see that the basis states corresponding to classical configurations , @xmath242 or @xmath243 , are most probable , although quantum corrections are already sizable . for longer chains
, the amplitude of the classical antiferromagnetic configuration continues to decrease , while quantum fluctuation contributions increase .
although a three - node chain is small , the analytic solution can serve as a starting point for understanding how an odd - size chain acts as an effective single spin . to study longer spin chains
, we numerically solve the eigenvalues and eigenvectors of the hamiltonian ( [ chain_hamil ] ) with .@xcite figure [ fig : odd_eng ] ( a ) plots a few lowest energy levels of a spin chain with @xmath80 nodes as a function of the external magnetic field @xmath16 .
when @xmath116 , the odd - size chain has two doubly degenerate ground states , with total magnetic quantum number @xmath123 , and denoted by @xmath244 and @xmath245 . an external magnetic field
@xmath16 splits the two degenerate ground states by the zeeman energy ( for small fields ) , labeled by @xmath58 .
thus , the odd - size spin chain can be considered as an effective single spin at small @xmath16 field and in the low energy limit , as shown in fig .
[ fig : odd_eng ] ( b ) . as a function of @xmath16 .
the two lowest levels around @xmath16 form up and down states of an effective single spin .
two gaps , @xmath58 and @xmath56 are indicated .
( b ) a magnified plot of the two lowest levels , offset against the ground state energy @xmath246 at @xmath102 , shows a zeeman splitting of a single spin @xmath1 . ]
two important energy scales for an odd - size chain are the energy gaps @xmath58 and @xmath56 .
the energy gap @xmath56 gives the splitting between the ground and excited manifolds .
the fact that @xmath58 is given by the zeeman splitting at low fields signifies that the doublet can be considered an effective spin-1/2 .
figure [ fig : odd_gap ] shows how the description of the odd - size chain as an effective single spin is limited by its size @xmath13 .
while the zeeman splitting @xmath58 is independent of the size @xmath13 of the spin chain around zero magnetic field , the range of the magnetic field @xmath16 , within which the effective spin picture is valid , does depend on @xmath13 .
the crossover behavior in @xmath58 occurs at smaller @xmath16 for larger @xmath13 because the manifold splitting , @xmath247 , decreases with @xmath13 . and @xmath56 measured in units of @xmath28 are plotted as a function of @xmath16 for @xmath248 . ]
now that we have established the fact that an odd - size spin chain can be considered as an effective spin-1/2 at low energy , we can construct an effective qubit - bus hamiltonian within the hilbert space spanned by the bus ground doublet states and the qubit states . with the doubly degenerate ground states @xmath249 of the odd - size bus ,
the projection operator @xmath224 is written as @xmath250 the effective hamiltonian to the first order in @xmath17 is @xmath251 where @xmath252 gives rise to a constant energy shift . in an external magnetic field
, it also gives the zeeman interaction of the effective spin-1/2 system of the spin bus or qubits , @xmath253 or @xmath254 with the external field . with @xmath116 , the first - order effective hamiltonian @xcite
is only given by @xmath255 , and takes the form @xmath256 where the effective coupling between qubit @xmath38 and the spin - bus @xmath2 is given by [ app : eq_coupling_j1 ] @xmath257 where @xmath20 is the bare coupling between the @xmath12th spin of the chain and the external qubit @xmath38 , and @xmath40 is the dimensionless local magnetic moment at the site @xmath12 of the chain in the ground state . in the case of @xmath240 , equation ( [ 3state ] ) gives the local magnetic moments , @xmath258 , with @xmath259 .
the effective hamiltonian ( [ app:1st_eff_hamil_odd ] ) shows again that an odd - size chain acts as an effective spin-1/2 particle that couples to the external qubits @xmath5 and @xmath0 , as illustrated in fig .
[ fig1 ] .
the effective hamiltonian to second - order in @xmath231 , which is needed for longer - time operations , is given by @xcite @xmath260 where the rkky - like second - order coupling @xmath4 is @xcite @xmath261 here @xmath45 are the energy levels and @xmath46 are the eigenstates of @xmath9 omitting the magnetic quantum number .
the summation with a prime indicates the exclusion of the ground states .
while the exact calculation of @xmath4 requires complete information of the eigenvalues and eigenstates of the chain , an approximate form can be obtained with _ only the ground state and the excitation gap_. using the closure relation @xmath262 , where @xmath237 sums through all the isolated - chain eigenstates , we obtain @xmath263 where @xmath49 refers to @xmath264 . in other words ,
the second - order coupling @xmath4 can be approximated by the difference between the spin - spin correlation function and a product of local magnetic moments of the ground state of the odd - size chain .
the competition between these two terms leads to decaying oscillations in @xmath265 .
note that a phase slip occurs when the qubit separation reaches a certain range , as discussed in ref . .
in an even - size chain with antiferromagnetic couplings , the spins are completely compensated , so that the chain has a non - degenerate ground state with both zero total magnetic moment , @xmath266 , and zero local magnetic moment , @xmath267 .
these properties of the even - size chain can be illustrated with the simplest case of @xmath268 , when the ground state is the singlet state @xmath269 . in fig .
[ fig : even_eng ] we plot the lowest energy levels of an even - size chain with @xmath210 as a function of the external magnetic field @xmath16 . for small @xmath16 , the non - degenerate ground state with @xmath217 is separated from the excited states by the gap @xmath58 .
similar to the odd - size chain , this gap and the field range before level crossing is limited by the size of the chain , as shown in fig .
[ fig : even_gap ] . as a function of @xmath16 .
the two lowest energy gaps , @xmath58 and @xmath56 in units of @xmath28 are indicated . ] and @xmath56 , of even - size chains with @xmath270 , as functions of @xmath16 . ]
protected by the @xmath58 gap , the low - energy hilbert space we are interested in is spanned by the ground state of the even - size chain and the qubit basis states .
the projection operator takes the form @xmath271 . the first order term in eq .
( [ eff_hamil ] ) is @xmath272 , which is just a constant energy shift . in other words , the first order perturbation term for an even - size chain
induces no qubit - bus couplings , since the bus ground state by itself can not have any dynamics .
the second - order perturbation term is obtained in the same way as in the case of an odd - size chain .
the effective hamiltonian to second order in @xmath231 is @xcite @xmath273 where the rkky - like coupling @xmath4 has the same form as eq .
( [ app : j_rkky ] ) , although the prime here would exclude only a single ground state .
similar to the case of the odd - size chain , @xmath4 can be approximated as @xmath274 where @xmath58 is the energy gap between the ground state and the first excited state , and @xmath275 is the spin - spin correlation function of the ground state of the chain .
there is no contribution from local magnetic moments here since they vanish in the ground state of an even - size bus .
here we prove that for an odd - size spin chain , the energy splitting @xmath58 of the ground doublet states at low uniform magnetic fields is independent of the randomness in the inter - node exchange coupling @xmath62 . with @xmath165 , the hamiltonian ( [ hamil_random ] ) is @xmath71 where @xmath72 is the unperturbed hamiltonian ( [ chain_hamil ] ) , and the perturbation @xmath73 is given by @xmath74 the two lowest eigenstates of the unperturbed hamiltonian @xmath72 are @xmath276 and @xmath277 , when the magnetic field is applied in the positive @xmath7 direction , and the corresponding eigenenergies are @xmath278 and @xmath279 .
for the perturbed hamiltonian @xmath9 , we denote the two lowest eigenstates by @xmath49 and @xmath142 , respectively , and the corresponding eigenvalues by @xmath84 and @xmath85 . by definition
, the energy gap @xmath58 of @xmath9 can be expressed as @xmath280 where @xmath281 is the lowest energy gap of the unperturbed hamiltonian @xmath72 , which is a zeeman gap .
the energy shifts of the two lowest levels , @xmath282 and @xmath283 , caused by the perturbation @xmath73 , i.e. , fluctuations @xmath66 , are given by @xcite @xmath284 their difference @xmath285 is thus @xmath286\,.\end{aligned}\ ] ] since @xmath64 = 0 $ ] , the two lowest states , @xmath49 and @xmath142 , ( also @xmath287 and @xmath288 ) of an odd - size bus are spin - flipped states of each other , so that the bracket part in the above equation vanishes , which leads to @xmath289 .
this means that the two lowest energy states fluctuate together , as shown in fig .
[ fig : odd_rj_gap ] ( a ) , so that their difference , i.e. , the zeeman energy gap , does not change .
we have thus proven that the zeeman energy splitting @xmath58 of the ground doublet states is invariant over random exchange couplings @xmath62 .
consequently , _ the odd - size chain with random exchange couplings can still be regarded as an effective single spin @xmath1 in the low energy limit . _
in this appendix we discuss the scaling properties of a spin chain as a quantum data bus .
while the heisenberg spin-1/2 chain is exactly solvable with bethe ansatz , @xcite only partial information about the ground state and the elementary excitations are available .
various numerical approaches have been applied to this system since bonner and fisher s pioneering work.@xcite although there has been tremendous advances in computational power , the exact diagonalization method can only handle a spin-@xmath1 system with sizes of up to @xmath290 , depending on the number of eigenvalues and eigenstates to be calculated .
indeed , one could consider such limitations as one of the motivations for building a quantum computer .
below we present our results for spin chains with @xmath13 up to 20 .
calculation of the first - order qubit - bus effective coupling of eq .
( [ app : eq_coupling_j1 ] ) needs only knowledge on the ground state of an odd - size chain . on the other hand , calculation of the second - order coupling ,
( [ app : j_rkky ] ) , requires the full knowledge of the eigenvalues and eigenstates of the spin chain .
we have solved the full spectrum of spin chains with sizes of up to @xmath291 on a personal computer using lapack @xcite , and obtained a few lowest energy eigenvalues and eigenstates for spin chains with sizes up to @xmath292 using primme .
@xcite per bond of the open chains ( open red circle ) and rings ( solid blue circle ) , ( b ) the gap @xmath293 between the ground state and the excited state , ( c ) the local magnetic moment @xmath294 at the end of the open - chain as a function of the size @xmath13 , and ( d ) a log - log plot of @xmath294 for odd - size chains .
the dashed line in panel ( a ) indicates @xmath295 obtained from the bethe ansatz @xcite .
the dashed line in panel ( b ) represents @xmath296 from ref . .
the gray lines in panels ( c ) and ( d ) represent an @xmath297 fit . ]
figure [ fig : scaling ] shows our results on the ground state energy @xmath84 per bond , the gap between the ground and the excited states , and the end - point local magnetic moment of the odd - size chain for finite chains of size @xmath268 to @xmath298 . as shown in fig .
[ fig : scaling ] ( a ) , the ground state energy per bond approaches the analytic value of @xmath299 in the thermodynamics limit , obtained from the bethe ansatz .
here the number @xmath300 of bonds is given by @xmath301 for open chains and @xmath302 for rings .
the ground state energy oscillates , depending on the even - odd parity , as the size of the chain @xmath13 increases . however , this finite - size effect diminishes in the large-@xmath13 limit .
figure [ fig : scaling ] ( b ) plots the ground state energy gap @xmath293 as a function of the size of the chain @xmath13 , with @xmath303 for even - size chains and @xmath304 for odd - size chains .
the numerical result follows the well - known analytical estimate @xmath305 as the size @xmath13 increases .
@xcite although the ground energy @xmath84 , the spin - spin correlation function for the ground state @xmath306 , and the ground energy gap @xmath293 are well known , @xcite the scaling property of the local magnetic moment of the odd - size chain is less well understood . @xcite figure [ fig : scaling ] ( c ) plots the dependence of the end - site local magnetic moment on the size @xmath13 .
our numerical data points to a @xmath297 dependence ( possibly slightly faster ) , as indicated in panel ( d ) of fig .
[ fig : scaling ] .
further work is still needed to clarify this @xmath13-dependence .
99 d. loss and d. p. divincenzo , * 57 * , 120 ( 1998 ) .
j. i. cirac and p. zoller , * 74 * , 4091 ( 1995 ) .
a. blais , r .- s .
huang , a. wallraff , s. m. girvin , and r. j. schoelkopf , * 69 * , 062320 ( 2004 ) .
s. bose , * 91 * , 207901 ( 2003 ) ; contem .
phys . * 48 * , 13 ( 2007 ) .
m. friesen , a. biswas , x. hu , and d. lidar , * 98 * , 230503 ( 2007 ) .
s. oh , m. friesen , and x. hu , * 82 * , 140403(r ) ( 2010 ) . y .-
shim , s. oh , x. hu , and m. friesen , , 180503 ( 2011 ) . s. oh , l .- a .
wu , y .- p .
shim , j. fei , m. friesen , and x. hu , , 022330 ( 2011 ) .
s. oh , * 65 * , 144526 ( 2002 ) ; x. hu and s. das sarma , , 012312 ( 2002 ) .
merkulov , a.l .
efros , and m. rosen , * 65 * , 205309 ( 2002 ) . m. a. ruderman and c. kittel , phys . rev . * 96 * , 99 ( 1954 ) ; t. kasuya , prog .
* 16 * , 45 ( 1956 ) ; k. yoshida , phys . rev . * 106 * , 893 ( 1957 ) . ed . by a. das and b. k. chakrabarti ( springer - verlag , berlin , 2005 ) .
d. scherrington and s. kirkpatrick , * 35 * , 1792 ( 1975 ) .
s. f. edwards and p. w. anderson , j. phys .
f * 5 * , 965 ( 1975 ) .
x. hu and s. das sarma , , 100501 ( 2006 ) .
p. fulde , _ electron correlations in molecules and solids _
( springer - verlag , berlin , 1991 ) . c. cohen - tannoudji ,
j. dupont - roc , and g. grynberg , _ atom - photon interactions _ ( john wiley & sons , new york , 1992 ) h. bethe , z. phys . * 71 * , 205 ( 1931 ) . c. f. hirjibehedin ,
c. p. lutz , and a. j. heinrich , science * 312 * , 1021 ( 2006 ) .
e. anderson , z bai , c. bischof , s. blackford , j. demmel , j. dongarra , j. du croz , a. greenbaum , s. hammarling , a. mckenney , and d. sorensen , _ users guide _
( siam , phiadelphia , 1999 ) . c. kittel ,
_ quantum theory of solids _ ( john wiley & sons , inc .
, new york , 1963 ) . y .-
( in preparation ) . j. c. bonner and m. e. fisher , phys . rev . * 135 * , a640 ( 1964 ) .
a. stathopoulos and j. r. mccombs , acm transaction on mathematical software , * 37 * , 21 ( 2011 ) .
e. lieb , t. schultz , and d. mattis , ann .
( n.y . ) * 16 * , 407 ( 1961 )
. w. j. caspers , _ spin systems _
( world scientific , singapore , 1989 ) .
w. nolting and a. ramakanth , _ quantum theory of magnetism _
( springer - verlag , heidelberg , 2009 ) . | a strongly coupled spin chain can mediate long - distance effective couplings or entanglement between remote qubits , and can be used as a quantum data bus .
we study how the fidelity of a spin-1/2 heisenberg chain as a spin bus is affected by static random exchange couplings and magnetic fields .
we find that , while non - uniform exchange couplings preserve the isotropy of the qubit effective couplings , they cause the energy levels , the eigenstates , and the magnitude of the couplings to vary locally . on the other hand , random local magnetic fields lead to an avoided level crossing for the bus ground state manifold , and cause the effective qubit couplings to be anisotropic .
interestingly , the total magnetic moment of the ground state of an odd - size bus may not be parallel to the average magnetic field .
its alignment depends on both the direction of the average field and the field distribution , in contrast with the ground state of a single spin which always aligns with the applied magnetic field to minimize the zeeman energy .
lastly , we calculate sensitivities of the spin bus to such local variations , which are potentially useful for evaluating decoherence when dynamical fluctuations in the exchange coupling or magnetic field are considered . |
on 2008 january 9.56 ut , while observing the supernova ( sn ) 2007uy in the nearby spiral galaxy ngc2770 ( @xmath0 ) , the x - ray telescope onboard _ swift _ detected a bright x - ray transient ( xt ) , with a peak luminosity of @xmath1 erg s@xmath2 and a duration of about 10 minutes @xcite .
its power - law spectrum and light curve shape were reminiscent of gamma - ray bursts ( grbs ) and x - ray flashes , but the energy release was at least 2 orders of magnitude lower than for typical and even subluminous grbs , also allowing for beaming ( e.g. , @xcite ) .
the discovery of the xt prompted the search for , and discovery of , an optical counterpart @xcite .
we performed spectroscopy of the source as soon as possible , starting 1.75 days after the xt , using the fors2 spectrograph on the eso very large telescope ( vlt ) .
subsequent spectroscopic monitoring of the object was carried out at the nordic optical telescope ( not ) and the william herschel telescope ( wht ) .
all spectra have been reduced using standard techniques . on january 18.22
ut ( 8.65 days after the xt ) we secured a high - resolution spectrum using the uves instrument on the vlt .
for this observation , we adopted the eso cpl pipeline ( v3.3.1 ) , and flux calibration was performed using the master response curves .
the observing log of the spectra is reported in table [ speclog ] .
imaging observations were conducted using the not , the vlt , the liverpool telescope ( lt ) and the united kingdom infrared telescope ( ukirt ) . image reduction was carried out using standard techniques . for photometric calibration , we observed optical standard star fields on five different nights , and defined a local sequence in the ngc2770 field . in the near - infrared we used two micron all sky survey stars as calibrators
magnitudes were computed using small apertures , and subtracting the background as measured in an annulus around the sn position .
the contribution from the underlying host galaxy light was always negligible , as also apparent from archival sloan digital sky survey images .
our photometric results are listed in table [ tab : photo ] .
one of our spectra covers the nucleus of ngc2770 , allowing a precise measurement of its redshift : @xmath3 .
this is slightly larger than the value listed in the nasa extragalactic database ( @xmath4 ) . for @xmath5 km s@xmath2 mpc@xmath2 ,
the luminosity distance is 29.9 mpc .
k , reddened assuming @xmath6 mag ( section [ sec : av ] ) .
the d narrow absorption from the interstellar medium in ngc2770 is also noted , as well as the two strong telluric features ( marked with ` @xmath7').[fg : earlyspec ] ]
our first optical spectrum of the transient source ( figure [ fg : earlyspec ] ) exhibits d absorption lines at @xmath8 , thus establishing its extragalactic nature .
broad features are also apparent across the whole spectrum ( @xmath9 km s@xmath2 ) , which led us to identify the object as a core - collapse sn @xcite .
@xcite describe nearly simultaneous spectra as featureless , probably due to their smaller covered wavelength range ( @xmath10 ) .
@xcite report features consistent with those in our data .
we initially classified the sn as a very young type ib / c , based on the absence of conspicuous si and h lines @xcite .
as the spectrum is among the earliest observed for any sn , comparable only to the very first spectrum of @xcite , there is no obvious resemblance with known sn spectra .
it is notable , however , that the earliest spectrum of the type - ic was essentially flat with broad , low - amplitude undulations ( though the covered wavelength range was limited ; @xcite ) .
early spectra , also mostly featureless , are available for the h - rich type - iip @xcite .
@xcite interpret them in terms of high temperature and ionization .
a striking feature in the spectrum is a conspicuous w - shaped absorption with minima at 3980 and 4190 ( rest frame ) .
it was detected using two different instrument setups ( figure [ fg : earlyspec ] ) , and also reported by @xcite . if interpreted as due to p cyg profiles , the inferred expansion velocity is @xmath11 km s@xmath2 , computed from the position of the bluest part compared to the peak .
its origin is unclear , although , following @xcite , @xcite propose that it is due to a combination of , , and .
interpreting the broad absorption at @xmath12 as @xmath13 6347 , 6371 , some ejecta reached @xmath14 km s@xmath2 .
such large velocities have been seen only in broad - lined ( bl ) type - ic sne , at significantly later stages @xcite . ]
figure [ fg : lc ] shows the optical and near - infrared light curves of . in the first days after the xt , the flux dropped faster in the bluer bands , with the color becoming progressively redder .
this can be interpreted as due to the stellar envelope cooling after the shock breakout @xcite .
our first spectra were taken during this stage , before energy deposition by radioactive nuclei became dominant , hence the physical conditions of the emitting material might be different than later .
we note that the w - shaped absorption discussed in section [ sec : earlyspec ] was no longer visible from 3.5 days after the xt onward ( fig .
[ fg : specevo ] ) .
the later spectra , acquired during the radioactivity - powered phase and extending over more than two months in time , established as a type - ib sn @xcite . from january 17 and
onward unambiguous he lines are observed ( figure [ fg : specevo ] ) , consistent with other reports @xcite . in figure
[ fg : photov ] , we plot the velocities at maximum absorption of a few transitions determined using the synow code @xcite . for comparison , we also plot the velocity of the bl @xcite and of the normal type - ic @xcite , showing that the velocities of are lower than those of bl sne .
stands for `` high - velocity h@xmath15 '' ) . the narrow emission line at 6560 is residual h@xmath15 from the sn host galaxy .
the feature around 6800 is affected by the b - band atmospheric absorption.[fg : specevo ] ] it follows from the detection of strong d with an equivalent width ( ew ) of 1.3 that the extinction toward is substantial in ngc2770 .
our best estimate of the reddening comes from comparing the colors of with those of stripped - envelope sne , which have @xmath16 around maximum ( e.g. , @xcite ) .
the resulting reddening is @xmath6 mag , corresponding to an extinction @xmath17 mag ( using the extinction law by @xcite with @xmath18 ) .
a large dust content is supported by absorption features in our high - resolution spectrum ( see also @xcite ) .
the d1 absorption line indicates a multicomponent system , spanning a velocity range of 43 km s@xmath2 , which sets a lower limit @xmath19 mag ( * ? ? ?
* their figure 4 ) .
the d versus @xmath20 relation for sne @xcite suggests @xmath21 mag .
diffuse interstellar bands ( dibs ) are also detected at 5781.2 , 5797.8 , and 6283.9 ( rest frame ) .
their ews suggest @xmath22 mag @xcite
. a dusty environment has been directly revealed through millimeter imaging of ngc2770 @xcite .
last , the hydrogen column density in the x - ray spectrum of the xt is @xmath23 @xmath24 ( assuming solar abundances ; @xcite ) .
the gas - to - dust ratio is @xmath25 @xmath24 mag@xmath2 , close to the galactic value @xmath26 @xmath24 mag@xmath2 @xcite .
( open symbols ) .
the latter interpretation is unlikely due to the lack of corresponding h@xmath27 ( see also @xcite ) .
shows velocities lower than the prototypical hypernova @xcite and comparable to @xcite.[fg : photov ] ]
the precise explosion epoch is so far only known for a few type - ii sne , thanks to either the detection of the neutrino signal (;
@xcite ) or of the uv flash by _ galaxy evolution explorer _ @xcite , and for bl type - ic sne associated with grbs ( e.g. , @xcite ) . is the first type - ib sn with a precisely constrained explosion epoch , since the xt is expected to occur less than 1 hr after the stellar collapse @xcite .
the nature of the xt shock breakout versus relativistic ejecta is still debated @xcite , thus it is unclear whether this phenomenon is common .
the light - curve evolution of is very similar to that of @xcite .
the initial fading can be interpreted as due to the envelope cooling through expansion following the initial x - ray / uv flash @xcite .
the subsequent rebrightening is due to the energy released by radioactive material in the inner layers and gradually reaching the optically thin photosphere . from our @xmath28 data
, we constructed the bolometric light curve of ( figure [ fg : lcbolo ] ) .
for comparison we also show the two other he - rich sne caught during the early cooling phase : @xcite and @xcite . strikingly , the three sne had very similar light curves during the photospheric phase .
given its peak luminosity , synthesized about @xmath29 of @xmath30ni based on arnett s rule @xcite .
the emission during the early cooling phase , however , varied substantially for the three sne ( seen only in @xmath31 for ) .
furthermore , significant radiation may be emitted blueward of the @xmath31 band during this early phase , so that the bolometric values may be underestimated .
has different properties from the sne associated with grbs , namely the presence of he in the ejecta , a lower peak luminosity ( @xmath32 mag ) , lower expansion velocities ( a factor of @xmath33 ) , and lower @xmath30ni mass ( a factor of @xmath34 ) .
the sn environment is also unlike that of grbs @xcite .
whether these two kinds of high - energy transients are separate phenomena or form a continuum is unclear .
addressing this issue will require theoretical modeling and an enlarged sample .
the discovery of a short - lived xt associated with an ordinary type - ib sn opens the possibility of accessing the very early phases of ordinary sne , which will provide new insights into sn physics .
future x - ray sky - scanning experiments , such as lobster or erosita , may turn out , rather unexpectedly , ideally suited to examine this issue , alerting us to the onset of many core collapse sne .
the dark cosmology centre is supported by the dnrf .
is supported by the spanish research programs aya2004 - 01515 and esp2005 - 07714-c03 - 03 and p.m.v . by the eu under a marie curie intra - european fellowship , contract meif - ct-2006 - 041363 .
p.j . acknowledges support by a marie curie european re - integration grant within the 7th european community framework program under contract number perg03-ga-2008 - 226653 , and a grant of excellence from the icelandic research fund .
we thank m. modjaz and m. tanaka for discussion , and the observers at vlt , not , wht , ukirt , in particular a. djupvik , s. niemi , a. somero , t. stanke , j. telting , h. uthas , and c. villforth .
amati , l. 2006 , , 372 , 233 arnett , w. d. 1982 , , 253 , 785 bionta , r. m. , et al .
1987 , , 58 , 1494 campana , s. , et al .
2006 , , 442 , 1008 cardelli , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 chevalier , r. , & fransson , c. 2008 , , 683 , l135 cox , n. l. j. , kaper , l. , foing , b. h. , & ehrenfreund , p. 2005
, , 438 , 187 deng , j. , & zhu , y. 2008 , gcn circ . 7160 dessart , l. , et al . 2008 , , 675 , 644 filippenko , a. v. , et al . 1995 , , 450 , l11 fisher , a. , branch , d. , hatano , k. , & baron , e. 1999 , , 304 , 67 folatelli , g. , et al . 2006 , , 641 , 1039 galama , t. j. , et al .
1998 , , 395 , 670 gezari , s. , et al .
2008 , , 683 , l131 ghirlanda , g. , nava , l. , ghisellini , g. , & firmani , c. 2007 , , 466 , 127 gorosabel , j. , et al . 2008 , l , arxiv:0810.4333 , submitted hjorth , j. , et al .
2003 , , 423 , 847 hirata , k. , et al .
1987 , , 58 , 1490 li , l .- x . 2007 , , 375 , 240 li , l .- x .
2008 , , 388 , 603 malesani , d. , et al .
2008 , gcn circ .
7169 mazzali , p. a. , et al .
2006 , , 442 , 1011 mazzali , p. a. , et al .
2008 , science , 321 , 1185 menzies , j. w. , et al .
1987 , , 227 , 39 modjaz , m. , li , w. , butler , n. , et al . 2008a , , submitted ( arxiv:0805.2201 ) modjaz , m. , chornock , r. , foley , r. j. , filippenko , a. v. , li , w. , & stringfellow , g. 2008b , cbet 1221 munari , u. , & zwitter , t. 1997 , , 318 , 269 patat , f. , et al .
2001 , , 555 , 900 predehl , p. , & schmitt , j. h. m. m. 1995 , , 293 , 889 quimby , r. m. , wheeler , j. c. , hflich , p. , akerlof , c. w. , brown , p. j. , & rykoff , e. s. 2006 , , 666 , 1093 quimby , r. m. , aldering , g. , wheeler , j. c. , hflich , p. , akerlof , c. w. , & rykoff , e. s. 2007 , , 668 , l99 richmond , m. w. , treffers , r. r. , filippenko , a. v. , paik , y. , leibundgut , b. , schulman , e. , & cox , c. v. 1994 , , 107 , 1022 richmond , m. w. , et al .
1996 , , 111 , 327 sauer , d. n. , mazzali , p. a. , deng , j. , valenti , s. , nomoto , k , & filippenko , a. v. 2006 , , 369 , 1939 schawinski , k. , et al .
2008 , science , 321 , 223 soderberg , a. m. , et al . 2008a , , 453 , 469 soderberg , a. m. , berger , e. , fox , d. , cucchiara , a. , rau , a. , ofek , e. , kasliwal , m. , & cenko , s. b. 2008b , gcn circ .
7165 stanek , k. z. , et al .
2003 , , 591 , l17 stritzinger , m. , et al .
2002 , , 124 , 2100 tanaka , m. , et al .
2008 , , arxiv:0807.1674 , in press thne , c. c. , michalowski , m. j , leloudas , g. , cox , n. l. j. , fynbo , j. p. u. , sollerman , j. , hjorth , j. , & vreeswijk , p. m. 2008 , , arxiv:0807.0473 , submitted turatto , m. , benetti , s. , & cappellaro , e. 2003 , from twilight to highlight : the physics of supernovae , ed .
w. hillebrandt & b. leibundgut ( berlin : springer ) , 200 valenti , s. , turatto , m. , navasardyan , h. , benetti , s. , & cappellaro , s. 2008a , gcn circ .
7163 valenti , s. , delia , v. , della valle , m. , benetti , s. , chincarini , g. , mazzali , p. a. , & antonelli , l. a. 2008b , gcn circ .
7221 waxman , e. , mszros , p. , & campana , s. 2007 , , 667 , 351 xu , d. , zou , y. c. , & fan , y. z. 2008 , , arxiv:0801.4325 , submitted @lrll@ jan 11.31 & 1.75 & vlt / fors2+g300v & @xmath35 + jan 11.32 & 1.76 & vlt / fors2+g600b & @xmath36 + jan 13.07 & 3.51 & not / alfosc+g4 & @xmath37 + jan 15.17 & 5.61 & not / alfosc+g4 & @xmath37 + jan 15.95 & 6.39 & not / alfosc+g4 & @xmath38 + jan 16.26 & 6.70 & not / alfosc+g4 & @xmath38 + jan 17.20 & 7.64 & not / alfosc+g4 & @xmath37 + jan 18.22 & 8.66 & vlt / uves+dic#1 & @xmath39 + jan 26.15 & 16.59 & wht / isis+r300b / r316r & @xmath40 + jan 29.01 & 19.45 & not / alfosc+g4 & @xmath37 + feb 01.02 & 53.46 & not / alfosc+g4 & @xmath41 + feb 04.05 & 56.49 & not / alfosc+g4 & @xmath41 + feb 18.15 & 70.59 & not / alfosc+g4 & @xmath37 + feb 25.88 & 78.32 & not / alfosc+g4 & @xmath42 + mar 02.18 & 112.62 & not / alfosc+g4 & @xmath37 + mar 18.87 & 129.31 & not / alfosc+g4 & @xmath43 @llccl@ jan 11.26528 & 01.70082 & @xmath31 & [email protected] & not+stancam + jan 13.01426 & 03.44980 & @xmath31 & [email protected] & not+alfosc + jan 15.22931 & 05.66485 & @xmath31 & [email protected] & not+alfosc + jan 16.03116 & 06.46670 & @xmath31 & [email protected] & not+alfosc + jan 17.25019 & 07.68573 & @xmath31 & [email protected] & not+alfosc + jan 18.00103 & 08.43657 & @xmath31 & [email protected] & not+stancam + feb 01.06841 & 22.50395 & @xmath31 & [email protected] & not+alfosc + feb 02.00892 & 23.44446 & @xmath31 & [email protected] & not+alfosc + feb 04.03992 & 25.47546 & @xmath31 & [email protected] & not+alfosc + feb 11.03650 & 32.47204 & @xmath31 & [email protected] & not+alfosc + jan 11.04261 & 01.47815 & @xmath45 & [email protected] & not+stancam + jan 11.26873 & 01.70427 & @xmath45 & [email protected] & not+stancam + jan 13.01980 & 03.45534 & @xmath45 & [email protected] & not+alfosc + jan 15.23748 & 05.67302 & @xmath45 & [email protected] & not+alfosc + jan 16.03701 & 06.47255 & @xmath45 & [email protected] & not+alfosc + jan 17.25617 & 07.69171 & @xmath45 & [email protected] & not+alfosc + jan 18.00862 & 08.44416 & @xmath45 & [email protected] & not+stancam + jan 25.90236 & 16.33790 & @xmath45 & [email protected] & lt+ratcam + jan 29.05104 & 19.48658 & @xmath45 & [email protected] & not+alfosc + jan 30.98094 & 21.41648 & @xmath45 & [email protected] & not+alfosc + feb 01.07844 & 22.51398 & @xmath45 & [email protected] & not+alfosc + feb 02.01195 & 23.44749 & @xmath45 & [email protected] & not+alfosc + feb 04.04304 & 25.47858 & @xmath45 & [email protected] & not+alfosc + feb 11.04021 & 32.47575 & @xmath45 & [email protected] & not+alfosc + feb 18.98034 & 40.41588 & @xmath45 & [email protected] & not+alfosc + mar 01.09311 & 51.52865 & @xmath45 & [email protected] & not+alfosc + mar 10.88157 & 61.31711 & @xmath45 & [email protected] & not+mosca + apr 02.93574 & 84.37128 & @xmath45 & [email protected] & not+mosca + jan 11.02120 & 01.45674 & @xmath46 & [email protected] & not+stancam + jan 11.27136 & 01.70690 & @xmath46 & [email protected] & not+stancam + jan 13.02744 & 03.46298 & @xmath46 & [email protected] & not+alfosc + jan 15.24242 & 05.67796 & @xmath46 & [email protected] & not+alfosc + jan 16.04183 & 06.47737 & @xmath46 & [email protected] & not+alfosc + jan 17.26096 & 07.69650 & @xmath46 & [email protected] & not+alfosc + jan 18.01366 & 08.44920 & @xmath46 & [email protected] & not+stancam + jan 20.02994 & 10.46548 & @xmath46 & [email protected] & not+stancam + jan 25.90873 & 16.34427 & @xmath46 & [email protected] & lt+ratcam + jan 29.05856 & 19.49410 & @xmath46 & [email protected] & not+alfosc + jan 30.19193 & 20.62747 & @xmath46 & [email protected] & not+alfosc + jan 30.97634 & 21.41188 & @xmath46 & [email protected] & not+alfosc + feb 01.08387 & 22.51941 & @xmath46 & [email protected] & not+alfosc + feb 02.01399 & 23.44953 & @xmath46 & [email protected] & not+alfosc + feb 04.04479 & 25.48033 & @xmath46 & [email protected] & not+alfosc + feb 11.04238 & 32.47792 & @xmath46 & [email protected] & not+alfosc + feb 18.98778 & 40.42332 & @xmath46 & [email protected] & not+alfosc + mar 01.09818 & 51.53372 & @xmath46 & [email protected] & not+alfosc + mar 10.88568 & 61.32122 & @xmath46 & [email protected] & not+mosca + mar 17.86086 & 68.29640 & @xmath46 & [email protected] & not+alfosc + apr 02.94851 & 84.38405 & @xmath46 & [email protected] & not+mosca + apr 16.95106 & 98.38660 & @xmath46 & [email protected] & not+stancam @llccl@ jan 11.00922 & 01.44476 & @xmath47 & [email protected] & not+stancam + jan 11.26190 & 01.69744 & @xmath47 & [email protected] & not+stancam + jan 11.30417 & 01.73971 & @xmath47 & [email protected] & vlt+fors2 + jan 13.02367 & 03.45921 & @xmath47 & [email protected] & not+alfosc + jan 15.24730 & 05.68284 & @xmath47 & [email protected] & not+alfosc + jan 16.04659 & 06.48213 & @xmath47 & [email protected] & not+alfosc + jan 17.26593 & 07.70147 & @xmath47 & [email protected] & not+alfosc + jan 18.01774 & 08.45328 & @xmath47 & [email protected] & not+stancam + jan 20.04708 & 10.48262 & @xmath47 & [email protected] & not+stancam + jan 29.06847 & 19.50401 & @xmath47 & [email protected] & not+alfosc + jan 30.21722 & 20.65276 & @xmath47 & [email protected] & not+alfosc + jan 30.97162 & 21.40716 & @xmath47 & [email protected] & not+alfosc + feb 01.08694 & 22.52248 & @xmath47 & [email protected] & not+alfosc + feb 02.01603 & 23.45157 & @xmath47 & [email protected] & not+alfosc + feb 04.04619 & 25.48173 & @xmath47 & [email protected] & not+alfosc + feb 11.04420 & 32.47974 & @xmath47 & [email protected] & not+alfosc + feb 18.99276 & 40.42830 & @xmath47 & [email protected] & not+alfosc + mar 01.10190 & 51.53744 & @xmath47 & [email protected] & not+alfosc + mar 02.12923 & 52.56477 & @xmath47 & [email protected] & not+alfosc + mar 10.88805 & 61.32359 & @xmath47 & [email protected] & not+mosca + mar 17.86648 & 68.30202 & @xmath47 & [email protected] & not+alfosc + apr 02.95326 & 84.38880 & @xmath47 & [email protected] & not+mosca + apr 16.94566 & 98.38120 & @xmath47 & [email protected] & not+stancam + jan 11.03211 & 01.46765 & @xmath48 & [email protected] & not+stancam + jan 11.27381 & 01.70935 & @xmath48 & [email protected] & not+stancam + jan 15.25279 & 05.68833 & @xmath48 & [email protected] & not+alfosc + jan 16.05135 & 06.48689 & @xmath48 & [email protected] & not+alfosc + jan 17.27084 & 07.70638 & @xmath48 & [email protected] & not+alfosc + jan 18.02196 & 08.45750 & @xmath48 & [email protected] & not+stancam + jan 20.05173 & 10.48727 & @xmath48 & [email protected] & not+stancam + jan 29.07600 & 19.51154 & @xmath48 & [email protected] & not+alfosc + jan 30.22517 & 20.66071 & @xmath48 & [email protected] & not+alfosc + jan 30.99911 & 21.43465 & @xmath48 & [email protected] & not+alfosc + feb 01.09031 & 22.52585 & @xmath48 & [email protected] & not+alfosc + feb 02.01851 & 23.45405 & @xmath48 & [email protected] & not+alfosc + feb 04.04764 & 25.48318 & @xmath48 & [email protected] & not+alfosc + feb 11.04589 & 32.48143 & @xmath48 & [email protected] & not+alfosc + feb 18.99647 & 40.43201 & @xmath48 & [email protected] & not+alfosc + mar 01.07881 & 51.51435 & @xmath48 & [email protected] & not+alfosc + mar 06.11299 & 56.54853 & @xmath48 & [email protected] & vlt+fors1 + mar 10.89042 & 61.32596 & @xmath48 & [email protected] & not+mosca + mar 17.87137 & 68.30691 & @xmath48 & [email protected] & not+alfosc + apr 02.95712 & 84.39266 & @xmath48 & [email protected] & not+mosca + apr 16.95661 & 98.39215 & @xmath48 & [email protected] & not+stancam + jan 12.555 & 02.99100 & @xmath49 & [email protected] & ukirt+ufti + jan 14.598 & 05.03400 & @xmath49 & [email protected] & ukirt+ufti + jan 15.457 & 05.89300 & @xmath49 & [email protected] & ukirt+ufti + jan 17.645 & 08.08100 & @xmath49 & [email protected] & ukirt+ufti + jan 21.004 & 11.44000 & @xmath49 & [email protected] & not+notcam + jan 23.576 & 14.01200 & @xmath49 & [email protected] & ukirt+ufti + feb 16.240 & 37.67600 & @xmath49 & [email protected] & ukirt+wfcam + feb 25.295 & 46.73100 & @xmath49 & [email protected] & ukirt+wfcam + mar 24.250 & 74.69000 & @xmath49 & [email protected] & ukirt+wfcam + jan 12.555 & 02.99100 & @xmath50 & [email protected] & ukirt+ufti + jan 14.598 & 05.03400 & @xmath50 & [email protected] & ukirt+ufti + jan 15.457 & 05.89300 & @xmath50 & [email protected] & ukirt+ufti + jan 17.645 & 08.08100 & @xmath50 & [email protected] & ukirt+ufti + jan 21.004 & 11.44000 & @xmath50 & [email protected] & not+notcam + jan 23.576 & 14.01200 & @xmath50 & [email protected] & ukirt+ufti + feb 16.240 & 37.67600 & @xmath50 & [email protected] & ukirt+wfcam + feb 25.295 & 46.73100 & @xmath50 & [email protected] & ukirt+wfcam + mar 24.250 & 74.69000 & @xmath50 & [email protected] & ukirt+wfcam + jan 12.555 & 02.99100 & @xmath51 & [email protected] & ukirt+ufti + jan 14.598 & 05.03400 & @xmath51 & [email protected] & ukirt+ufti + jan 15.457 & 05.89300 & @xmath51 & [email protected] & ukirt+ufti + jan 17.645 & 08.08100 & @xmath51 & [email protected] & ukirt+ufti + jan 21.004 & 11.44000 & @xmath51 & [email protected] & not+notcam + jan 23.576 & 14.01200 & @xmath51 & [email protected] & ukirt+ufti + feb 16.240 & 37.67600 & @xmath51 & [email protected] & ukirt+wfcam + feb 25.295 & 46.73100 & @xmath51 & [email protected] & ukirt+wfcam + mar 24.250 & 74.69000 & @xmath51 & [email protected] & ukirt+wfcam | was discovered while following up an unusually bright x - ray transient ( xt ) in the nearby spiral galaxy ngc2770 .
we present early optical spectra ( obtained 1.75 days after the xt ) which allowed the first identification of the object as a supernova ( sn ) at redshift @xmath0 .
these spectra were acquired during the initial declining phase of the light curve , likely produced in the stellar envelope cooling after shock breakout , and rarely observed .
they exhibit a rather flat spectral energy distribution with broad undulations , and a strong , w - shaped feature with minima at 3980 and 4190 ( rest frame ) .
we also present extensive spectroscopy and photometry of the sn during the subsequent photospheric phase . unlike sne associated with gamma - ray bursts , displayed prominent he features and is therefore of type ib . |
the emergence of the world wide web ( www ) belongs arguably to the most relevant events of the present time .
the interest in this system and in networks in general permeated through all the society , including physics .
this led , at the turn of the century , to a large amount of studies of what with the time came to be known as `` network science '' .
most studies of the www were performed , however , in the early 2000s @xcite and large - scale studies of the www are rather hard to find nowadays , despite the immense growth of the internet in the last 10 years .
a remarkable finding of the first generation studies of the www is the emergence of scale - free degree distributions , which can be explained potentially from the view of preferential attachment , although the exponents obtained are not universal @xcite .
generally , one can assume that the growth process of a complex network will be influenced by inter - node correlations and that these dependencies will be reflected in the resulting network topology .
however , such correlations are not easy to detect and characterize , and have not been studied in depth .
it is expected that a simple rule as preferential attachment can not completely reproduce the structures found in real - world networks , and therefore more complicated models have been developed to replicate the behavior @xcite .
correlations between different properties are generally used as a proxy to study the internal structure of the network .
for instance , vespagnani studied correlations between the in - degree of a node and that of a first neighbor of said node @xcite , showing a scale free property ( recently modeled by takagi @xcite ) , barabasi and albert studied the local clustering coefficient as a function of the in - degree @xcite , in order to obtain information regarding the hierarchical structuring of the network . however , real - world data about said correlations is not abundant . in the present work we study the complete dominant core of the www by crawling @xmath0 million domains , including all domains with the largest traffic ( all domains with an alexa rank of one million or less are included ) .
collapsing the data , by neglecting link multiplicities , we study the network of inter - domain hyperlinks ( not webpages ) , containing about half a billion directed edges .
we find non - trivial correlations between in- and out - degree , between the in - degree and the local clustering coefficient and between the degrees of neighboring sites .
in addition to evaluate averaged quantities , we study the full density plots , finding novel scaling features for the boundaries of several correlation functions .
we present , in addition , a formula for the clustering coefficient of random directed graph characterized by given arbitrary in- and out - degree sequences .
finally we present an analysis of the correlations between the number of in - links and the alexa rank of a domain . and out - degree @xmath2 .
right : two types of in - degree clusters , with the edges always directed towards the central site ( a).,title="fig:",scaledwidth=11.0% ] and out - degree @xmath2 .
right : two types of in - degree clusters , with the edges always directed towards the central site ( a).,title="fig:",scaledwidth=11.0% ]
for directed graphs we have to distinguish between the distribution @xmath3 and @xmath4 of the in- and the out - degrees @xmath5 and @xmath6 , respectively . there are , in addition , two kinds of nearest neighbors , in - neighbors and out - neighbors .
site b1 in fig .
[ fig : graph - illustrations ] is a nearest in - neighbor of site a and a nearest out - neighbor of site b2 .
alternatively one could call b1 an ancestor of a and a descendent of b2 @xcite . for bi - directional links , as between b3 and b4 in fig .
[ fig : graph - illustrations ] , in - neighbors are also out - neighbors .
the total number of in - links equals the total number of out - links , the in- and out - degree coordination numbers @xmath7 are hence identical . in order to calculate the relevance of correlations between in- and out- degree in the structure of the network , we have developed a statistical model of the clustering coefficient for given distributions of in- and out- degree which are uncorrelated .
we define with @xmath8 the excess distribution @xcite of outgoing links of a nearest in - neighbor .
the normalization constant @xmath9 is just the coordination number @xmath10 , see ( [ def_z ] ) .
equivalently we define via @xmath11 the degree distribution ( not excess ) of incoming links of a nearest in - neighboring site . here
@xmath12 is the probability that a site has @xmath6 out - links and @xmath5 in - links ( joint distribution function ) , with the usual relations @xmath13 for the marginal distribution functions .
the normalization constant @xmath14 in ( [ def_bar - q - in ] ) is given by the coordination number @xmath10 , @xmath15 for the clustering coefficient @xmath16 ( the ` hat ' symbol stands here for the clustering coefficient of a directed graph ) we now consider two in - neighbors , having respectively , with probabilities @xmath17 and @xmath18 , @xmath5 in - links and @xmath6 excess out - links ( as stubs ) .
we now assume that the distributions @xmath19 and @xmath20 of the two neighbors are independent of each other .
the probability , for a graph with @xmath21 nodes , that a given pair of in- and out - stubs are connected is then @xmath22 , where @xmath23 is the total number of in- or out - stubs , and hence @xmath24 transforming now into a sum @xmath25 over sites , every site @xmath26 being characterized by an in - degree @xmath27 and out - degree @xmath28 , one obtains @xmath29 which coincides with the usual expression @xcite for non - directed graphs ( apart from a factor @xmath28 instead of @xmath30 in the first factor ) , by taking @xmath31 for @xmath32 . a fully - connected network results in @xmath33 under this formula .
we note that the expression ( [ eq : hat - c - sites ] ) for @xmath16 may actually violate the sum rule @xmath34 , due to the neglect of inter - site degree correlations , when applied to a real - world graph . as an example consider a network composed out of a single star , like the site c in fig .
[ fig : graph - illustrations ] , but with bi - directional edges . for a un - directed ( and loopless ) star
the degree sequence is @xmath35 with an intensive coordination number @xmath36 . the statistical formula ( [ eq : hat - c - sites ] ) for the clustering coefficient would , one the other hand , diverge @xmath37 in the thermodynamic limit @xmath38 .
a substantial deviation of @xmath16 from the true clustering coefficient is hence a measure for the strength of inter - site degree correlations , the expression ( [ eq : hat - c - sites ] ) being valid for graphs with vanishing inter - node correlations . for the in - degree ( main panel ) and the out - degree ( inset ) , log - log plot .
the dashed lines , corresponding to power - law distributions , have slopes @xmath39 and @xmath40 , respectively for the in- and the out - degree.,scaledwidth=45.0% ]
using the crawlers of the former file search engine findfiles.net @xcite we crawled , mostly in 2011 , @xmath0 million domains ( of type http://www.domain.com ) with a total of @xmath41 million subdomains ( of type http://subdomain.domain.com ) .
these @xmath0 million domains have 223 million hyperlinks in between them , linking in addition to 50 million other sites . for the network analysis we neglected these 50 million external sites , as we did not crawl them separately .
the network of 223 million inter - domain directed links has an average degree of 32 and @xmath42 million of the @xmath0 million domains are isolated in the sense that they have no in - links , they can not be reached from the core of the world wide web .
a further one million sites have just a single hyperlink directed to them .
the crawling strategy started from the set of the about 32 million subdomains referred - to in wikipedia and dmoz ( all languages ) , with further systematic additional extensions .
we included , in particular , the one million domains with the largest traffic volume , in terms of the alexa rank .
this data set , which we denote with ff-2011 , hence corresponds essentially to the complete relevant part of the world wide web , in terms of traffic volume .
the degree distribution of hyperlinks have been observed to follow a power law @xmath43 , with an exponent close to the limiting case @xmath44 ( when the mean degree would diverge in the thermodynamic limit ) @xcite . in fig .
[ fig : in - out - degree - distribution ] we present the complementary cumulative distribution functions @xcite for both the in - degree and the out - degree . over a range of about 2.5 - 3 orders of magnitude
, the data can be approximated quite nicely by power law distributions , with exponents @xmath45 and @xmath46 respectively for the in- and the out - degree .
these results confirm earlier studies @xcite finding consistently @xmath47 .
the absolute magnitude of the values reported for the scaling exponents vary slightly from study to study , either because of the evolution of the internet with time passing , or due to the size of the respective databases . of domains with in - degree @xmath5 and out - degree @xmath48 .
the density is shown in log scale , as well as both axes .
the solid line represents the average @xmath49 .
the probability density is normalized , @xmath50.,scaledwidth=45.0% ] in fig .
[ fig : kvskout ] , the density distribution of nodes having an in - degree @xmath5 and an out - degree @xmath48 , is presented , together with the average out - degree @xmath51 , for sites having an in - degree @xmath5 . in- and out - degree do not seem to be particularly correlated , on a first sight .
however , the average out - degree shows two regimes with approximated power - law scaling , for @xmath52 and @xmath53 , with exponents @xmath54 and @xmath55 respectively . in the case that the joint distribution @xmath56 would factorize , @xmath57 , the mean out - degree @xmath58 would functionally follow the in - degree distribution @xmath3 , where @xmath59 is the average ( in- and out- ) degree of our internet data .
however , as shown in fig .
[ fig : in - out - degree - distribution ] , the marginal in - degree distribution @xmath3 , falls approximately like @xmath60 , viz substantially faster than ( [ eq : in_out ] ) would imply .
in- and out - degree are hence non - trivially correlated .
we will discuss the nature of the respective correlations in more detail further below when discussing the distribution of local clustering coefficients .
the local clustering coefficient @xmath61 is given by the number of linked nearest neighbors of site @xmath62 , relative to the total number of possible links between the neighbors .
for directed graphs there are in- and out - neighbors and various possible 3-site loops , as illustrated in fig .
[ fig : graph - illustrations ] , also known as network motifs @xcite . here
we examine the in - clustering coefficient . for a given site @xmath62 the in - clustering coefficient @xmath61
is given by the average number of links in between the in - neighbors of site @xmath62 . in fig .
[ fig : graph - illustrations ] , the sites ( b1,a , b2 ) form an in - loop of site a , contributing to @xmath63 , while the sites ( b4,a , b3 ) contribute two in - loops .
we focus on the in - clustering coefficient since the number of in - links is a measure for the importance of a site , contributing to its traffic volume .
we find , for the ff-2011 network data , a mean clustering coefficient @xmath64 of @xmath65 .
this is , for two reasons , a surprising high value .
firstly the connection probability @xmath66 is very low , being just @xmath67 .
secondly a quite large number of sites , 0.27% , has a vanishing local clustering @xmath68 , and only a small fraction , 0.3% , of domains , mostly with small degrees , have a maximal local clustering coefficient of unity .
we can assess the impact of correlations on the formation of local loops by considering identical degree sequences for the in- and out- degree , as extracted from the ff-2011 network data , but considering various types of correlations between the in- and out- degree of each node .
* applying eq .
( [ eq : hat - c - sites ] ) to the actual network , the obtained value amounts to @xmath69 .
this value is over the maximum @xmath70 , indicating towards very strong correlations between the in- and the out - degree distributions , compare the discussion below eq .
( [ eq : hat - c - sites ] ) . * for a network having the same degree distributions @xmath3 and @xmath4 for the in- and the out - degree as the actual network , but without correlations between these degrees , viz assuming a joint probability distribution @xmath71 , the clustering coefficient obtained by eq .
( [ eq : hat - c - sites ] ) would amount to @xmath72 . *
a network where the in- and out- degrees are anticorrelated ( nodes with largest in - degree are mapped to the smallest out - degree ) , would amount to an even lower @xmath73 .
* for a network with a maximally correlated distribution of in- and out- degree ( nodes with the largest in - degree being mapped to the largest out - degree ) , would result again in a higher - than - unity clustering coefficient @xmath74 , when using eq .
( [ eq : hat - c - sites ] ) .
we hence conclude that the in- and out - degree are quite strongly correlated positively for the world wide web . of pairs of in - degree
@xmath5 and local clustering coefficient @xmath75 .
the density , and both axes are given in log scales .
the solid line represents the average value @xmath76 , as a function of in - degree @xmath5 .
the probability density is normalized , @xmath77.,scaledwidth=45.0% ] in fig .
[ fig : kvsc ] , the density of @xmath78 pairs is shown in a log - log scale , where @xmath5 is the in - degree and @xmath75 the local clustering coefficient .
the density distribution has upper and lower cutoffs scaling approximatively like @xmath79 , with @xmath80 and @xmath81 .
the lower limit has a simple explanation .
the lowest non - zero local clustering coefficient is realized when just a single loop exist out of the @xmath82 possible triangles , @xmath83 when setting the number of loops @xmath84 to one .
the exponent of the upper limit , @xmath85 , implies , compare ( [ eq : c_min ] ) , that the number of local loops scales like @xmath86 .
we have presently no explanation for this scaling behavior . the average value of @xmath87 , as a function of in - degree @xmath5 , follows mostly a power law for small @xmath88 , with an exponent @xmath89 . for
larger @xmath90 the exponent changes toward @xmath91 for the mean local clustering coefficient .
this last exponent is in agreement with previous observations found in @xcite , and are a fingerprint for a hierarchical network structure .
the change in behavior at the point @xmath92 is also observable in the correlation between the in - degree and the degree of nearest neighbors , as we will show in the next sections .
there is a group of nodes with very high clustering coefficients @xmath93 around the @xmath94 region ( close to where the upper limit with the @xmath95 slope intersects the abscissa ) , which somewhat falls of the line . after analyzing some of the domains involved in this region ,
we conclude that this group of nodes does not represent the intrinsic network structure of the www , belonging most probably to _ link farms_. these nodes are however responsible for the jumps in @xmath87 at @xmath94 . of the in - degree @xmath5 and the in - degree @xmath96 of nearest neighbor sites .
both the density distribution and the axis are given in log scale .
the solid line corresponds to the average value @xmath97 .
the probability distribution is normalized , such that @xmath98.,scaledwidth=45.0% ] in fig .
[ fig : kvsknn ] , the density of pairs @xmath99 is shown , where @xmath96 is the in - degree of a first neighbor , and @xmath5 the in - degree .
the dots with higher densities for low @xmath5 values are relatively large groups of linked domains which share exactly the same pairs of in - degree and first - neighbor in - degree .
the domains do not seem to be particularly related although we do not discard the possibility that they may belong to _ link farms _ , as they clearly stand out of the general behavior of the density distribution .
when analyzing the average @xmath100 as a function of the in - degree @xmath5 , we observe a very weak increase for small @xmath5 until @xmath101 .
we can fit this increase fairly good with a power law of exponent @xmath102 .
this behavior would be in agreement with the one observed in the canonical barabasi - albert model @xcite , though it differs with with a 1998 www network study @xcite . in the range from @xmath101 to @xmath103
we observe a change in the behavior of the average @xmath100 , as it starts decaying with increasing @xmath5 .
this decay follows a power - law as well , with an exponent of about @xmath104 .
this decay is closer to the results found in @xcite for a subset of the 1998 internet data and the fitness model developed therein ( which decays with @xmath105 ) .
however , the decay is observed in our results for much higher degree @xmath5 than in @xcite , which has data limited to @xmath106 .
we speculate that this difference is due to the size of the network studied , although it might be possible that the evolution of the www in the last 10 years is responsible for the structural change . , where @xmath107 is the alexa index and @xmath5 the corresponding in - degree of the domain .
the density is given in log scale , as well as both axes .
the solid line shows the average @xmath108 , as a function of @xmath107 .
the probability is normalized , such that @xmath109.,scaledwidth=45.0% ] we have analyzed the correlations of the alexa rank @xcite with respect to the in - degree @xmath5 .
the alexa rank is arguably one of the most popular measures of the traffic received by an internet site , and so , its relevance .
the ranking is proprietary , so the general public does not have access to the specifics of its calculation , although according to the official information , it is derived from the traffic observed , with data partly retrieved from users who installed the alexa add - on to their web browser @xcite . in this ranking
, the site with the most traffic has rank @xmath110 , the following largest rank @xmath111 and so on
. the rank does not provide any information about the precise amount of traffic , such that a larger @xmath107 index does not give any indication of how much less traffic does that site receives , but rather only that it receives less traffic than the sites with smaller @xmath107 . in fig .
[ fig : alexavsk ] , we present the density of domains as a function of its in - degree @xmath5 and alexa rank @xmath107 .
we only analyze the alexa rank for sites having an in - degree @xmath112 , with a few exceptions , due to constraints in retrieving the alexa rank data .
we observe a distribution limited from above and below by two power laws with exponents @xmath113 and @xmath114 .
the lower limit is however less pronounced , due to the lack of samples . the solid line in fig .
[ fig : alexavsk ] shows the average in - degree @xmath108 , for sites having and alexa rank @xmath107 .
we observe a very marked power law decay @xmath115 the exponents are not the inverse of each other , since @xmath116 and @xmath117 are distinct averages .
there is a saturation at @xmath118 for the scaling regime , presumable due to our constraint @xmath112 for the alexa index .
we find this scaling particularly interesting , since the alexa rank is not derived directly from the topology of the network but rather from the traffic generated by users
. for site administrators the relatively weak decay ( [ eq : scaling_k - a ] ) implies that the traffic generated by in - coming hyperlinks can be a relevant contribution to the overall traffic volume .
we have performed a visualization study of the data discussed hitherto by considering relative joint distribution functions , which are obtained by dividing a given joint probability distribution by the product of the respective marginal distributions , @xmath119 in the absence of correlations , viz when @xmath120 , the respective density plots would be homogeneous and flat . in fig .
[ fig : correl]a , we show the relative density @xmath121 for pairs of in - degree @xmath5 and local clustering coefficient @xmath75 .
the distribution is quite homogeneous and the upper and lower limits of the distribution are exalted in comparison with the plot of the bar joint distribution presented in fig .
[ fig : kvsc ] . in fig .
[ fig : correl]b , we present the relative joint density for the correlation between in- and out - degree . the distribution is considerably more homogeneous than the respective bare probability density shown in fig .
[ fig : kvskout ] .
however , a substantial enhancement remains for small in- and out- degrees . from the shape of the joint nearest - neighbor degree distribution presented in in fig .
[ fig : kvsknn ] , it would be tempting to think that its shape is mostly determined by the marginal distributions , i.e. that @xmath122 , and that therefore @xmath5 and @xmath96 would be essentially decorrelated .
this is , however , not the case , as we can see in fig .
[ fig : correl]c . in this plot
the correlations are seen more clearly in terms of the relative joint distribution we can observe that the resulting relative distribution still shows a stronger correlation when both in - degree and the in - degree of the neighbor are small . in fig .
[ fig : correl]d , we present the relative density distribution of for sites having an alexa rank @xmath107 and an in - degree @xmath5 . as for the case of the clustering coefficient @xmath75 ,
the distribution maintains the very marked upper and lower limits , being otherwise essentially only slightly more uniform than the orginal data shown in fig .
[ fig : alexavsk ] .
the present study shows that many properties of the www are characterized by non - trivial correlations .
we observe that the joint probability distributions , for several of the properties tested , follow power - law scaling for the respective averages . additionally , the distributions have , in many instances , density distributions which are limited by power laws .
the power law limiting functions do not share exponents neither with that of the marginal distributions , nor with the respective average value of the property studied .
interestingly one also observes power - law scaling for a seemingly unrelated quantity , the alexa traffic rank , which decays as a function of in - degree and also the average in - degree decays moderately weakly as a function of the alexa rank .
we found that the statistical properties of the world wide web differ remarkably for domains receiving more / fewer than about @xmath1 hyperlinks from different domains .
the change in behavior is observed for the correlations between in- and out - degree , between in - degree and local clustering coefficient and between in - degree and the in - degree of neighbors .
this observation points towards an underlying hierarchical structure of the www , with the `` elite '' of the internet domains , receiving links from more than one thousand different domains , being made - up by about @xmath123 sites .
we would like to acknowledge the support of the german science foundation ( dfg ) . | we performed a large - scale crawl of the world wide web , covering @xmath0 million domains , including all high - traffic sites of the internet .
we present a study of the correlations found between quantities measuring the structural relevance of each node in the network ( the in- and out - degree , the local clustering coefficient , the first - neighbor in - degree and the alexa rank ) .
we find that some of these properties show strong correlation effects and that the dependencies occurring out of these correlations follow power laws not only for the averages , but also for the boundaries of the respective density distributions .
in addition , these scale - free limits do not follow the same exponents as the corresponding averages . in our study
we retain the directionality of the hyperlinks and develop a statistical estimate for the clustering coefficient of directed graphs .
we include in our study the correlations between the in - degree and the alexa traffic rank , a popular index for the traffic volume , finding non - trivial power - law correlations .
we find that sites with more / less than about @xmath1 links from different domains have remarkably different statistical properties , for all correlation functions studied , indicating towards an underlying hierarchical structure of the world wide web . |
the idea of constructing nanometer - sized devices and machines has created a lot of excitement during the last years . despite the progress made ,
the functionality of artificial nano - devices is , however , still rather limited . at the same time , more and more biomolecular nano - machines have been identified in the cells of living beings where they accomplish a huge variety of tasks .
many of these molecular motors are now rather well studied and were found to work with an amazing degree of precision and efficiency as a result of billions of years of evolution , @xcite . in the following , we will focus on one class of molecular motors which has been studied quite extensively during the last decade , namely processive cytoskeletal motors which drive the traffic of vesicles and organelles within cells .
these motors hydrolyze adenosinetriphosphate ( atp ) and convert the free energy from this chemical reaction into directed movements along filaments of the cytoskeleton .
this class of motors contains kinesins and dyneins , which move along microtubules , and certain myosins , which move along actin filaments .
these motors walk along the filaments by performing discrete steps with a step size which corresponds to the repeat distance of the filament , 8 nm for kinesins and 36 nm for myosin v. they are called processive if they make many steps while staying in contact with the filament . from a physical point of view , much of the interest in molecular motors is due to the fact that the difference in size compared to macroscopic engines implies also conceptual differences .
the typical energy of macroscopic motors is much larger than the thermal energy , @xmath0 , while the typical energies of molecular motors are of the order of @xmath1 .
for example , the hydrolysis of atp releases about 20 @xmath1 . on the one hand , molecular motors have to cope with perturbations arising from thermal fluctuations ; on the other hand , the unavoidable presence of noise suggests that evolution has created motors which make use of this noise in order to generate work or directed movement . from a technological viewpoint ,
the amazing properties of single biological motor molecules and the complexity of the systems into which they are integrated in the cell provide inspiration for the design of artificial nanoscale transport systems @xcite . in this article , we discuss several theoretical aspects of the motor movements . in section [ sec : ratchets ] , we start by summarizing some recent experimental results and discuss the question whether noise - driven mechanisms are used by these motors . in section
[ sec : rw ] , we discuss another effect of noise , namely the detachment of motors from their tracks due to thermal fluctuations , which leads to peculiar random walks .
we derive the asymptotic behavior of these random walks using the statistical properties of the returns of motors to the filament . finally , in section [
sec : traffic ] , we summarize our recent studies of traffic problems which arise in systems with many molecular motors due to their mutual exclusion from binding sites of the filaments .
these topics are also addressed in a recent longer review article @xcite .
molecular motors can be studied outside cells using biomimetic model systems . in these experiments ,
the biological complexity is reduced to a minimal number of components , namely motors , filaments , and atp .
an example is shown in fig .
[ fig : beadassay ] . by these experiments
, one can observe movements of single motor molecules and measure transport properties such as velocities , step sizes , and forces @xcite .
on the one hand , these experiments provide insight into the motor mechanisms .
a major breakthrough was to resolve the discrete steps of the motors and to measure the step size of the motor which corresponds to the repeat distance of the filament .
this has first been achieved for kinesin using an optical tweezers setup @xcite .
more recently , it has been shown that kinesin @xcite and also myosin v @xcite move in a hand - over - hand way , i.e. , that the two heads of the dimer step forward in an alternating fashion such that the rear head always moves in front of the leading head , similar to human walking .
further progress is expected from combining mechanical methods and particle tracking with fluorescence techniques .
on the other hand , using these biomimetic motility assays , one can measure the transport properties systematically varying external control parameters . here
the main focus has been on the velocity as a function of the atp concentration and of the force applied with , e.g. , optical tweezers to oppose the movements , see , e.g. , ref .
other quantities that have been measured are the one - dimensional diffusion coefficient of motors bound to filaments or the randomness parameter and the walking distance before unbinding from the filament .
these measurements have stimulated a large amount of theoretical work , see , e.g. refs .
@xcite , modeling the walks of motors along filaments in order either to fit the experimental data or to find out the generic properties of these walks .
for example , it turns out that the motor velocity as a function of the atp concentration is given by a universal relationship which should be valid for many types of motors @xcite . since nanometer - sized molecular motors have to live in a noisy environment , it has soon been speculated whether these motors exploit the noise to generate their directed movements , and various variants of ratchet models have been proposed as reviewed in refs . @xcite . in the simplest case ,
the conformational changes associated with the chemical cycle of the motor rectify the one - dimensional brownian motion along the filament .
while such a simple mechanism is not consistent with the measurements for dimeric motors such as conventional kinesin or myosin v , it can describe movements of processive monomeric motors such as the monomeric kinesin kif1a @xcite ( for other processive monomeric motors , see the review @xcite ) .
these monomeric motors exhibit biased , but strongly diffusive movements along the filaments similar to what one obtains in the simplest ratchet models .
interestingly , coupling two such ratchets by a spring @xcite leads to a driving mechanism which is independent of the diffusion along the filament . similarly ,
dimerization of monomeric kinesin results in a higher velocity and smaller diffusion coefficient of the filament - bound motors @xcite .
these results suggest that the dimerization of motors has two effects : it allows the motor to stay bound to the filament for many chemical cycles in contrast to many ( therefore unprocessive ) monomeric motors and in addition , it allows the motors to move by a more efficient mechanism which , in contrast to the motility of monomers , does not rely on diffusion along the filament .
the fact that nanometer - sized motor molecules work in a noisy environment has another consequence : even processive motors do not move along a filament forever , but unbind from it after a certain binding time ( which corresponds to a typical walking distance ) , because the binding energy is finite and can be overcome by thermal fluctuations . for a single kinesin or myosin v motor ,
the binding times and walking distances are of the order of 1 s and 1 @xmath2 m , respectively .
much longer walking distances can be obtained if several motors form a complex or if a cargo is transported by a larger number of motors .
note that the unbinding may also have a biological function by allowing the motors to diffuse around obstacles on the filament .
unbound motors perform simple brownian motion until they rebind to the same or another filament . on large time scales ,
the combination of active directed movements along filaments and non - directed brownian motion leads to peculiar random walks , called motor walks in the following , which consist of alternating sequences of the two types of movements as sketched in fig .
[ fig : randwalk ] @xcite .
in order to determine the effective transport properties of these motor walks , we have studied several simple arrangements of filaments embedded in compartments of various geometries .
a particularly simple but intriguing case consists of a single filament and a set of confining walls , which restrict the diffusion of unbound motors . in the simplest case ,
there are no confining walls and the unbound motors can diffuse freely in the full three - dimensional space ( similar behavior is obtained for a half space geometry which is more easily accessible to experiments ) . by placing the filament in a quasi two - dimensional slab or in a cylindrical tube ( geometries , which are also accessible to _ in vitro _
experiments ) , diffusion can be restricted along one or two dimensions perpendicular to the filament .
we have studied the motor walks by mapping them to random walks on a lattice @xcite .
a line of lattice sites represents the filament .
motors at these sites perform a biased random walk and move predominantly into one direction , which we choose to be the positive @xmath3 direction . in a discrete - time description ,
they move with a small probability @xmath4 to each of the adjacent non - filament sites and thus unbind from the filament . at the non - filament sites the motors perform simple symmetric random walks and
move to each neighbor site with probability @xmath5 ( @xmath6 denotes the spatial dimension ) and rebind to the filament with probability @xmath7 when they reach again a filament site .
confining walls are implemented as repulsive boundaries , at which all attempted movements towards the walls are rejected .
we have used scaling arguments , computer simulations , and exact solutions of the master equations to study the drift and diffusion behavior arising from the motor walks @xcite .
the motor walks exhibit anomalous drift behavior and strongly enhanced diffusion parallel to the filament due to the repeated binding and unbinding . at large times , motors move with an effective velocity given by @xmath8 , where @xmath9 is the velocity of the bound motor and @xmath10 is the probability that the motor is bound to the filament . in the tube geometry
, @xmath10 is time - independent at large times and given by the equilibrium of binding / unbinding and diffusion perpendicular to the filament . in the slab and half - space geometries as well as in two- and three - dimensional systems without confining walls , a steady state can not be reached at any time , because motors can rebind to the filament after arbitrarily large excursions , and the typical size of these excursions , that contribute to the average behavior , increases with time @xmath11 .
therefore , @xmath10 and the effective velocity are time - dependent in these cases , namely @xmath12 for compartments with @xmath13 dimensions of unconfined diffusion ( for @xmath6-dimensional systems without confining walls , we have @xmath14 ) .
the time - dependent effective velocity implies that the average displacement of the motors grows sublinearly . in the effectively two - dimensional slab geometry ( as well as on a two - dimensional lattice without confining walls )
, the displacement behaves as @xmath15 at large times , and in the half space ( or full three - dimensional space ) , it is given by @xmath16 , see fig .
[ fig : displacement ] . in the following
, we will give a simplified description which explains these features and relates them to known results from the theory of random walks .
for simplicity , we discuss the case of a single infinitely long filament embedded into @xmath6-dimensional space with @xmath17 or @xmath18 .
motor particles binding to this filament walk along it with velocity @xmath9 until they unbind .
unbinding occurs with a rate @xmath19 , so that the motors perform straight movements over the walking distance @xmath20 .
the diffusive excursions between two bound walks bring the motors back to the filament .
( for simplicity , we take the sticking probability @xmath7 to be one . if @xmath21 , the motor has to return to the filament @xmath22 times before rebinding . )
the distribution @xmath23 of the excursion times @xmath24 is therefore given by the distribution of the return times of a random walker s return to a line in @xmath6 dimensions or , if we consider the projection into the plane perpendicular to the filament , of the return to the origin in @xmath25 dimensions .
this is a classical problem in the theory of random walks which was solved by polya in 1921 and has lead to the remarkable result that the return of a random walker to the origin is certain on one- and two - dimensional lattices , but not in three dimensions @xcite . for the molecular motors
, this implies that they will return to the filament with certainty .
if we are only interested in the movement parallel to the filament , we can consider the excursions away from the filament as periods of rest .
in addition , for large times , the duration of an excursion is typically much longer than the time the motor is bound to the filament ; therefore we can consider the walks along the filament as effectively instantaneous steps of size @xmath26 ( if @xmath21 , the effective step size is @xmath27 .
the motor walks are then described by continuous time random walks with the dwell time distribution @xmath23 given by the distribution of the excursion times . for this type of random walks
, solutions can be obtained using fourier
laplace transforms @xcite
. a short summary of this method is given in appendix [ sec : app ] . in that way
, one finds that the laplace transforms of the first two moments of the position of this random walker are given by @xmath28}\ ] ] and @xmath29}+\frac{2\langle x_s^2\rangle \psi^2(s)}{s[1-\psi(s)]^2}\ ] ] where @xmath30 is the laplace transform of the waiting time distribution @xmath31 . in general ,
@xmath32 and @xmath33 are the moments of the step size distribution ; for the motor walks , they are given by the active walks along the filament and we take them to be given by @xmath34 and @xmath35 , respectively . from these relations , the average displacement and the dispersion of
the motor walks for large times can be obtained by inverting the laplace transform .
in particular , these relations imply that the asymptotic displacement of the motors is given by the large - time ( or small @xmath36 ) behavior of the distribution of return times to the origin in @xmath25 dimensions .
normal drift behavior with @xmath37 is obtained , as long as @xmath23 has a finite mean value , @xmath38 , and thus @xmath39 for small @xmath36 .
if , however , @xmath23 decays slower than @xmath40 at large @xmath24 , the mean dwell time diverges ( which implies a divergence of @xmath41/s$ ] for small @xmath36 ) , and anomalous drift is obtained .
the latter behavior occurs in our case ( where the waiting times are given by return times to the filament ) if the diffusion away from the filament is not restricted . in the two - dimensional case
, the return time distribution behaves as @xmath42 for large @xmath24 or @xmath43 for small @xmath36 as shown in appendix [ sec : app : return ] . inserting this into eq .
( [ asmoments_psiabh ] ) and inverting the laplace transform , we obtain @xmath44 and @xmath45 for small @xmath36 and large @xmath11 , respectively . similarly , in the three - dimensional case , the return time distribution is @xmath46 for large @xmath24 or @xmath47 for small @xmath36 , see again appendix [ sec : app : return ] , which leads to @xmath48 likewise , we can obtain the dispersion @xmath49 of the motors from the second moment of the distribution arising from the encounters with filaments .
( [ asmoments_psiabh2 ] ) leads to @xmath50 and @xmath51 in two and three dimensions , respectively , where we have added the contribution due to the diffusion of unbound motors parallel to the filament . note that the broadening of the distribution of motors due to the encounters with the filament is characterized by an anomalously high effective diffusion coefficient of the order of @xmath52 in two dimensions , while in three dimensions , the leading term is the unbound diffusion , but with a large logarithmic correction , again of the order @xmath52 .
these results agree with the corresponding asymptotic results from the exact solution of the full master equations @xcite .
finally , we consider systems with many interacting molecular motors . the simplest type of motor motor interaction is simple exclusion which arises from the fact that a motor occupies a certain volume and , in particular , excludes other motors from the binding site of the filament to which it is bound as observed in decoration experiments ,
see , e.g. , ref .
exclusion is most important if motors accumulate in certain regions along the filaments , where it leads to the formation of molecular traffic jams .
these interactions are easily incorporated into our lattice model by rejecting all movements to occupied lattice sites @xcite .
these models then represent new variants of exclusion processes or driven lattice gas models , where the active or driven movements are localized to the filaments . in driven exclusion processes , the state of the system depends crucially on the boundary conditions , since the boundaries determine the current through the system .
we have studied tube systems with a single filament located along the axis of a cylindrical tube and with several types of boundary conditions at the left and right tube ends @xcite .
we use the convention that the active movement along the filament is biased to the right .
there are several possibilities to build such tube systems artificially such as micropipette glass tubes or liquid microchannels , but there are also several tubular compartments within cells for which these tube systems provide simple descriptions .
the most prominent example for the latter is the axon of a nerve cell ; another example is provided by the hyphae of fungi .
the simplest situation is given by _
periodic boundary conditions _ which can be solved exactly @xcite . in this case
, motor particles reaching the right end of the tube simply restart their movements at the left end .
this leads to constant density profiles for both the bound and unbound motors in the stationary state .
diffusive currents in the radial direction vanish and binding to the filament is locally balanced by unbinding .
the current of motors through the tube is given by @xmath53 and the value of the bound density @xmath54 is determined by the total number of motor particles in the tube . if the number of motors within the tube is increased beyond an optimal number ( where @xmath55 and @xmath56 ) , the current through the tube decreases due to jamming of the motors . in a _ closed tube _ , motors accumulate in front of the right end , and a diffusive current of unbound motors to the left balances the current along the filament in the stationary state @xcite .
if the number of motors in the tube is small , the motors are essentially localized at the right tube end . upon increasing the number of motors within the tube ,
a jammed region at the right tube end builds up , separated from a low density region to its left by a rather sharp interface , which provides probably the simplest example for active pattern formation by molecular motors .
the crowded domain spreads to the left at higher motor concentrations until the filament is uniformly covered by motors and rather crowded , see fig .
[ fig : tube ] .
such density profiles have recently been observed for a kinesin - like motor in fungal hyphae @xcite .
let us note that these density patterns exist due to continuous consumption of atp . if the atp concentration is not kept constant within the closed compartment , then after burning all atp , the density pattern will finally become homogeneous . along the filament to the right and
build up a motor traffic jam at the right end of the system . in the stationary state ,
the bound motor current is balanced by a current of unbound motors to the left arising from the diffusive movement of the unbound motors which is characterized by the diffusion coefficient @xmath57 .
( b ) profiles of the bound motor density as a function of the coordinate @xmath3 along the filament for various total numbers @xmath58 of motors within the tube .
the jammed region becomes longer with increasing @xmath58 . ]
tubes with _
open boundaries _ which are coupled to reservoirs of motors at both ends , so that motors enter the tube at the left end and leave it at the right end , exhibit boundary - induced phase transitions @xcite .
there are three phases which are distinguished by the bottleneck which determines the motor current through the tube .
this bottleneck can be the left boundary , the right boundary or the interior of the tube .
these three cases corresponds to the low - density ( ld ) , high - density ( hd ) and maximal - current ( mc ) phase as shown in fig . [ fig : pd_ot ] .
if changing the motor densities in the reservoirs at the boundaries leads to a change in the bottleneck position , a phase transition occurs which can be either discontinuous ( ld hd ) or continuous ( ld mc and hd
. these types of phases and transitions are known from the one - dimensional asymmetric simple exclusion process ( asep ) @xcite which corresponds to the dynamics along the filament in our model without the binding and unbinding processes .
the presence of the unbound state of the motors , however , increases the number of possible boundary conditions , and since the phase transitions are boundary - induced , the location of the transition lines within the phase diagram is quite sensitive to that choice . at present , it is difficult to see which of these boundary conditions will be the simplest to implement experimentally .
nevertheless , systems of molecular motors are promising candidates for the experimental observation of boundary - induced phase transitions .
another type of phase transitions occurs in systems with two - way traffic of motors : while each motor moves either towards the plus- or towards the minus - end of the corresponding filament , different types of motors move into opposite directions .
we have studied systems with two species of motors moving into opposite directions along the same filament @xcite .
the interactions of these motors are described by a single interaction parameter @xmath59 by which the binding and unbinding rates are increased or reduced in order to enhance binding and reduce unbinding if a motor of the same type is present at a neighbor site and to enhance unbinding and reduce binding if a motor with opposite directionality is bound at a neighbor site .
this type of interaction is suggested by decoration experiments .
for equal concentrations of both motor species and for sufficiently strong interaction with @xmath60 ( where @xmath61 is the critical value of the interaction parameter which depends on the overall motor concentration ) , there is spontaneous symmetry breaking , so that one motor species occupies the filament , while the other one is largely excluded from it , see fig .
[ fig:2spec ] .
if several filament are aligned in parallel and with the same orientation , this symmetry breaking leads to the spontaneous formation of traffic lanes for motor traffic with opposite directionality .
varying the relative concentration of the two motor species for @xmath60 leads to a discontinuous phase transition with hysteresis , similar to the transitions induced by changes of the external fields in magnetic systems .
in contrast to the boundary - induced phase transitions discussed in the preceding section , these transitions do not depend on the boundaries , but are induced by the binding and unbinding dynamics .
this implies that they are hardly affected by the choice of the boundary conditions .
they depend , however , on the active movements of the motors , and are not found in an equilibrium situation with motor velocity @xmath62 , as applies , e.g. , to systems without atp .
in summary , molecular motors exhibit interesting movements on several length scales . here we have addressed two of these length scales , namely the walks along filaments which consist typically of @xmath63 100 steps with a step size of the order of 10 nm and random walks which consist of many such walks along filament interrupted by periods of diffusion after unbinding from the filament .
in addition , the presence of many motors leads to traffic phenomena such as traffic jams and traffic lanes , similar to the macroscopic traffic on streets .
however , unbinding of motors from the filaments due to thermal fluctuations ( which is a consequence of their microscopic size ) plays an important role and can help to circumvent obstacles and to regulate the traffic .
in this appendix , we summarize some results for random walks with a dwell time distribution @xmath23 and a step distribution @xmath64 which are used in section [ sec : rw ] .
we consider random walkers in one dimension which make the first step at time @xmath65 starting from the origin , @xmath66 .
the probability distribution @xmath67 of such a random walk fulfills the recursion relation @xcite @xmath68 where @xmath69 is the probability density that the walker is at position @xmath3 after the @xmath70th step , @xmath71 is the probability that the @xmath70th step occurs at time @xmath11 , and @xmath72 is the probability that no step occurred until time @xmath11
. the initial conditions are @xmath73 and @xmath74 .
the solution of this recursion can be obtained using fourier
laplace transforms , which leads to @xmath75}\ ] ] for the fourier
laplace transform of the probability distribution @xmath67 @xcite with the laplace transform of the waiting time distribution , @xmath31 and the fourier transform of the step distribution @xmath76 .
the latter expansion is valid for small @xmath59 , provided that the moments @xmath77 of the step distribution @xmath64 with @xmath78 are finite . using this expansion
, we can derive expressions for the laplace transforms of the moments of our random walk by expanding @xmath79 as given in eq .
( [ ctrw_sq ] ) in powers of @xmath59 @xcite which leads to eq .
( [ asmoments_psiabh ] ) from which the asymptotic behavior of the time - dependent moments can by obtained via the tauberian theorems , see ref . @xcite .
in this appendix we sketch the derivation of the distribution of return times to the origin which determines the dwell time distribution @xmath80 used in section [ sec : asymptotics ] .
we consider the probability @xmath81 that , at time @xmath11 the random walker is at position @xmath82 for _ the first time _ provided it started at the origin at time @xmath65 .
@xmath83 then determines the distribution of return times to the origin .
these probabilities satisfy the recursion relation @xmath84 which states that the walker is at site @xmath82 at time @xmath11 , if it had been there at any time @xmath85 and returned there in time @xmath86 .
the last term expresses the initial conditions . using fourier
laplace transforms , one can derive an expression for the laplace transform of @xmath83 , the probability that the walker returns to the origin at time @xmath11 @xcite , @xmath87 where the @xmath88 are integrals over the momentum @xmath89 which depend on the spatial dimensions .
the same integrals have to be calculated for the exact solution of the motor walks @xcite . in the one - dimensional case we have @xmath90 which leads to @xmath91 for small @xmath36 and @xmath92 for large @xmath11 . in the two - dimensional case
the corresponding integral is given by @xmath93 where @xmath94 is a complete elliptic integral of the first kind and @xmath95 .
since @xmath94 behaves as @xmath96 for @xmath97 close to one @xcite , the return time distribution is asymptotically given by @xmath98 for small @xmath36 or @xmath99 for large @xmath11 .
the dwell time distribution @xmath23 used in section [ sec : asymptotics ] for the case of a filament within a @xmath6-dimensional lattice is given by @xmath100 , where the expression on the right hand side has to be taken in the @xmath14 dimensions . | processive molecular motors which drive the traffic of organelles in cells move in a directed way along cytoskeletal filaments . on large time scales , they perform motor walks , i.e. , peculiar random walks which arise from the repeated unbinding from and rebinding to filaments .
unbound motors perform brownian motion in the surrounding fluid .
in addition , the traffic of molecular motors exhibits many cooperative phenomena .
in particular , it faces similar problems as the traffic on streets such as the occurrence of traffic jams and the coordination of ( two - way ) traffic .
these issues are studied here theoretically using lattice models . , and molecular motors , active movement , random walks , lattice models , traffic jams 87.16.nn , 05.40.-a , 05.60.-k |
feedback from active galactic nuclei ( agn ) is a key ingredient in modern galaxy evolution models .
it is required in order to explain the sharp drop - off in the galaxy mass function above @xmath0 , prevent the cooling catastrophe in galaxy clusters and produce the hot gas atmospheres seen around many galaxies . observations of massive kpc - scale outflows and pc - scale relativistic winds provide further evidence that agn affect their host galaxies in a significant way .
one outstanding issue in understanding the precise effects of feedback is the range of spatial scales over which it operates .
the agn jets and winds are launched from the accretion disc on scales @xmath1 pc , while the observed effects on host galaxies span @xmath2 pc or more .
this means that in order to model the agn feedback precisely , a simulation should span at least 5 orders of magnitude in linear scale , resulting in @xmath3 resolution elements for a 3d model , a resolution which is not likely to be reached any time soon .
even resolving the interaction of an agn wind with the host galaxy material , on scales of several parsecs and larger , is currently only possible in single - galaxy models , rather than cosmological simulations @xcite . as a result of this shortcoming
, some effects of agn feedback might not be apparent from simulations .
in a recent paper @xcite , we showed that mass resolution has a significant influence upon the agn feedback effects seen in simulations of turbulent gas .
in particular , at mass resolutions typical of cosmological simulations , we found agn feedback to be much more negative , in the sense of efficiency of gas removal , than at higher resolution .
in high - resolution simulations , agn winds expel only diffuse gas , while denser gas forms filaments which fall in toward the central supermassive black hole ( smbh ) and can potentially feed it or form stars .
these results suggest that agn feedback can be positive as well as negative , increasing the star formation rate in its host .
our other previous research @xcite provides some support for this conclusion as well .
several other authors have investigated similar positive agn feedback coming from the interaction of the jet with the galactic ism @xcite . in this paper
, we investigate another numerical aspect of modelling agn feedback : the prescription detailing how the agn feedback energy is passed to the surrounding gas .
we perform high - resolution simulations of agn feedback acting upon a turbulent gas reservoir , passing the feedback energy in one of two ways : a monte carlo radiative transfer scheme , which self - consistently takes into account gas optical depth ; and a kernel - weighted thermal and kinetic energy input scheme .
we find that the monte carlo scheme , while being numerically far more expensive and difficult to scale , provides qualitatively different results than thermal and kinetic energy injection .
in particular , the monte carlo scheme allows the formation of simultaneous gas inflows and outflows , resulting in prolonged agn feeding and faster , but less massive , outflows .
we also test an adjusted version of both schemes , where the feedback energy is injected , or virtual particles released , biconically , rather than spherically symmetrically .
this change makes the results of the two schemes significantly more alike .
we suggest that using this improvement to the thermal feedback prescriptions would make the results of large - scale numerical simulations much more realistic .
the structure of the paper is as follows . in section [ sec : nummodel ] , we describe the numerical model and setup used in the simulations .
the simulation results are presented in section [ sec : results ] .
we discuss the implications of our work for both numerical modelling of galaxy evolution and interpretation of observations in section [ sec : discuss ] and conclude in section [ sec : concl ] .
the code used is gadget-3 , an updated version of the publicly available gadget-2 @xcite .
it is a hybrid n - body / sph code with individual particle timesteps and adaptive smoothing .
we employ the sphs extension of the basic sph scheme @xcite , which improves the modelling of mixing within a multiphase medium , allowing us to better track the interaction between turbulent gas flows .
we use a wendland kernel @xcite with 100 neighbours .
the initial conditions of all models presented in this paper are the same .
they consist of a gas sphere with isothermal density profile , extending from @xmath4 pc to @xmath5 kpc .
the gas sphere is resolved with @xmath6 particles , giving a particle mass of @xmath7 and a mass resolution of @xmath8 .
the spatial resolution of the model varies depending on gas density , but is generally of order a few pc ; we adopt a minimum gravitational and sph smoothing length of @xmath9 pc .
the gas is embedded in a static isothermal background density profile with velocity dispersion @xmath10 km / s , and is supported against the background gravity by turbulent motion with characteristic velocity @xmath11 .
the gas is initially cold ( @xmath12 ; see below for a discussion on how the gas evolves thermally ) .
a smbh is embedded in the centre of the gas distribution and can swallow gas which comes within the accretion radius @xmath13 pc of the smbh particle and is energetically bound to it .
we set the smbh particle mass to @xmath14 , in order for the radius of its sphere of influence to be @xmath15 .
this ensures that the gas moves in a purely isothermal background potential only perturbed by its own self - gravity , rather than the gravity of the smbh .
the smbh is inactive for the first @xmath16 myr of the simulation ; this allows the gas to develop a turbulent density structure before it is affected by feedback . in order to prevent the formation of spurious high - density filaments close to the inactive smbh , we extend the accretion radius to @xmath17 during this time .
after @xmath16 myr , the agn switches on for @xmath16 myr and begins affecting the gas in two ways .
first of all , we model the effect of the agn radiation field by employing a @xcite cooling curve , appropriate for optically thin gas subjected to agn radiation . below @xmath18
k , we extend the cooling curve with the one proposed by @xcite down to a temperature floor , which depends on the gas density such that the jeans mass of the gas never falls below the resolution limit .
gas particles which fall on this dynamic temperature floor are stochastically converted into sink particles in order to speed up the simulations .
this is equivalent to having a density threshold @xmath19 @xmath20 where @xmath21 k. this crude approximation of the star formation process may affect our results slightly , because sink particles are affected only by gravity and cease to interact with the agn feedback .
since sink particles are created in dense gas , this slightly reduces the fraction of feedback energy received by dense gas and increases the fraction received by diffuse gas . on the other hand , we do not model the effects of self - shielding of dense gas clumps , so the reduction in received energy caused by the formation of sink particles mimics that effect to some extent . in any case , this effect is small , since our simulations never produce more than @xmath22 sink particles . the other way that the agn affects the gas is by means of wind feedback , which we describe below .
[ cols="^,^,^,^,^,^,^",options="header " , ] the agn wind feedback model @xcite has a number of appealing properties : it is based upon relatively well - understood physical processes ; it can explain the @xmath23 relation without requiring free parameters ; it can explain the properties of observed fast agn winds on sub - parsec scales , as well as those of massive agn outflows on kpc scales @xcite .
observations show that winds with velocity @xmath24 and power @xmath25 are present in a large fraction of agn , suggesting a wide opening angle @xcite . in the wind feedback model ,
these winds shock against the galactic ism and drive large - scale outflows @xcite .
the shocked wind passes its energy and momentum to the ism , accelerating the outflow to @xmath26 km / s @xcite .
however , a large fraction of the outflow energy can leak out through gaps in the non - uniform ism , leading to a situation where some of the gas is outflowing with large velocities , while dense gas is simultaneously inflowing toward the smbh @xcite . therefore , it is important to check how well numerical simulations can capture this complex process of simultaneous inflow and outflow on scales from tens to thousands of parsecs .
we do this by exploring two subgrid feedback implementations and their improvements .
the simplest feedback prescription , often used in cosmological models , is the `` thermal feedback '' subgrid prescription . in this prescription ,
the feedback energy of the agn , which we take to be @xmath27 of its luminous energy output over the smbh timestep , is passed as thermal energy to the smbh particle neighbours .
we choose 100 neighbours for the smbh particle , the same as for gas particle .
testing showed that using fewer neighbours results in feedback becoming extremely inefficient , as gas particles can be accreted by the smbh even after being heated ; however , if the particles were not accreted , it is likely that the feedback would be more efficient as the temperature would be higher and so the cooling time would increase . meanwhile , using more neighbours dilutes the feedback and results in faster cooling of gas @xcite . the feedback energy given to each particle
is weighted by the sph interpolation kernel .
this means that gas closer to the smbh receives proportionately more energy than gas further away .
in addition to thermal energy , each sph particle receives a similarly kernel - weighted fraction of the agn wind momentum , @xmath28 , in a direction radially away from the smbh .
while this is not usually done in cosmological models , we implement this aspect of feedback in order to be able to do a more direct comparison with the monte carlo radiative transfer method .
tests showed that the effect of momentum feedback is negligible compared with the thermal energy input .
one issue with the simple thermal feedback prescription is that the feedback energy is given to the nearest particles without taking their spatial distribution into account .
this implicitly assumes that the gas distribution is ( approximately ) spherically symmetric .
a real galaxy contains multiphase gas , with many structures degrading the spherical symmetry , even if such symmetry might be expected on large scales . for the agn , this results in neighbouring gas particles having very different density and correspondingly very different effective cross - sections . in principle , feedback affecting each particle should be attenuated by calculating the optical depth between the smbh and this particle .
however , doing so is a very time - consuming process , unfeasible to implement in cosmological simulations , which may have many agn at the same time .
therefore , we propose a simpler method , which we believe captures some of the important physics .
the method relies on injecting the thermal and kinetic energy in a cone centered on the smbh instead of spherically symmetrically .
spherical symmetry is unlikely on the sub - grid scales on which the wind is launched , since the agn has a preferred plane - that of the accretion disc - and winds are likely to be launched perpendicular to it .
so we would expect the wind to be launched with an axisymmetric geometry .
this prescription ensures that gas in all direction does not experience identical feedback .
instead , there is a preferred plane for the gas to accrete in , corresponding to the plane of the sub - grid accretion disc .
numerically , we implement this model by calculating a separate neighbour list for the smbh particle , looking for neighbours only in a bicone with a specified opening angle , which is set to 45 degrees in the simulations presented in this paper . for simplicity ,
we choose the axis of the cone to lie in the z direction .
the number of neighbours in the cone is the same as the number of neighbours used for the hydrodynamics calculations , and the feedback energy input is weighted by the value of the sph interpolation kernel , only using the larger smoothing length corresponding to the neighbours in the cone .
the `` virtual particle '' monte carlo radiative transfer method @xcite relies on packets of feedback energy , which are emitted by the agn as particles and move in straight lines with a constant velocity , carrying momentum and energy .
when the virtual particle moves into an sph particle s smoothing kernel , it starts giving up its momentum and energy over several timesteps , distributing it proportionately among all sph particles which contain the virtual particle in their smoothing kernels . since the virtual particles are emitted isotropically , the energy injection is also isotropic , independent of the distribution of gas around the smbh .
this ensures that , for example , a dense and small gas clump close to the smbh does not receive too much of the feedback energy , because the prescription self - consistently takes into account the cross section of the clump .
it also means that even gas very far away from the smbh may receive a direct energy and momentum input , if the intervening space is empty , for example having been cleared by a progressing outflow .
we test two versions of the virtual particle model : one with spherically symmetric emission of tracer particles , and one where the particles are emitted in a cone with a 45-degree opening angle .
we present the results of 13 models ; their identification labels and main parameters are given in table [ table : param ] .
one simulation , noagn , is intended as a control and does not have any agn feedback ( the smbh is still present and acretes gas ) .
then four sets of three models each investigate the effects of different subgrid feedback prescriptions on feedback from agn of different luminosities .
we choose @xmath29 to span a range of observed quasar luminosities ; furthermore , these luminosities encompass the critical luminosity necessary to drive away gas from the agn by purely momentum feedback @xcite .
the opening angle of the cone in the thermal conical ( tkc ) and virtual particle conical ( vpc ) simulations is @xmath30 degrees .
at first , we briefly discuss the properties of the gas distribution just before the agn switches on
. then we present the morphology of the spherically symmetric feedback models , vp and tk , and finally move on to their conical feedback counterparts , vpc and tkc .
later , we discuss the resulting gas morphology , outflow energetics , inflow and outflow rates and the density structure . during the first 1 myr of evolution , the gas shell develops a turbulent density structure and
cools down to temperatures in the range @xmath31 k. figure [ fig : init_phase ] shows the phase diagram of the gas just before the agn switches on .
most of the gas has temperatures just below @xmath22 k , where there is a minimum in the cooling rate .
some gas is located on the temperature floor ( diagonal line in the bottom right of the distribution ) , however the probabilistic conversion of gas into sink particles ensures that only @xmath32 particles have been converted into stars .
approximately @xmath33 , i.e. @xmath34 , of the initial shell has been accreted by the smbh particle .
this makes the smbh mass grow to more than three times its initial mass , however this does not affect the subsequent results , since we fix the agn luminosity , rather than tying it to the smbh mass .
it is important to note that during this phase , the accretion radius is set to @xmath35 pc , and most of the gas accreted through this radius is unlikely to reach the smbh during the time period simulated in our models .
figure [ fig : vp_morph ] shows the density maps of the three virtual particle simulations at @xmath36 myr , i.e. @xmath37 myr after the agn switches on , showing the xz plane .
the plots use a wedge - slice projection , such that only particles with @xmath38 are plotted , to ensure that both the innermost regions and the outskirts are well - sampled .
the three plots reveal a clear change in the global effect of feedback . in the vp - l1 model ,
feedback is not strong enough to prevent inflow or even to produce significant bubbles ; there is only one small cavity in the negative - x direction .
simulation vp - l2 shows noticeable outflow bubbles , but simultaneously there exists gas very close to the smbh , in fact accreting upon the smbh particle .
most of the feedback energy leaks out in the directions of low density , producing these large bubbles , while the remaining momentum push is too weak to shut off accretion of dense gas .
finally , in vp - l5 , we see that feedback is powerful enough to blow away almost all the material , except for a couple of very dense filaments ; these remain close to the smbh , but are also pushed away , albeit slowly . in figure
[ fig : tk_morph ] , we plot density maps of the analogous tk models , again at at @xmath36 myr .
there are significant differences from the vp models .
most noticeably , a well - defined almost spherical feedback bubble is present in simulation tk - l5 , without any inflow toward the smbh . on the other hand ,
the size of the outflow bubble is much smaller in tk - l5 than in vp - l5 , since the energy is injected into the gas closest to the smbh ; this gas is dense and therefore can not be removed as efficiently as the diffuse gas which absorbs most of the input energy in the vp models .
it is also interesting to note that the feedback bubble is surrounded by a boundary layer of dense gas , which is not present in the virtual particle simulations .
this dense gas cools down very rapidly , so most of the feedback energy is radiated away rather than used to drive bubble expansion .
the major difference between the two models arises because in the thermal feedback simulation , all of the feedback energy is always injected into gas that is closest to the smbh , typically within the central @xmath39 pc , although the precise radius depends on agn luminosity and the size of the forming bubble . in the virtual particle simulations , on the other hand , feedback evacuates cavities which allow the virtual particles to stream to large distances of order @xmath40 pc in the highest - luminosity simulation .
we now present the effects of changing feedback geometry to conical , which ensures that there is always a channel - the midplane of the gas distribution - through which gas can reach the smbh , and similarly that there is a channel perpendicular to the first through which feedback can be injected at distances comparable to those in the spherical virtual particle model .
figure [ fig : vpc_morph ] shows the gas density in the three conical virtual particle models , vpc - l1 ( left ) , vpc - l2 ( middle ) and vpc - l5 ( right ) . even the lowest - luminosity simulation produces a modest outflow , with cavities in the top half and third quadrant of the plot , where the gas was least dense at the time the agn switched on .
the outflow is somewhat conical , but variations in gas density prevent it from acquiring a well - defined shape . in the highest - luminosity model , the feedback bubbles are rather large , touching the edges of the initial gas distribution , but gas close to the midplane has not been blown away , allowing filamentary inflow to continue .
it is worth noting that the outflow is wider than the opening angle of the cone within which the virtual particles are emitted .
this happens because the low - density cavities are over - pressurized and expand both vertically and laterally .
figure [ fig : tkc_morph ] shows the analogous snapshots of the tkc simulations tkc - l1 ( left ) , tkc - l2 ( middle ) and tkc - l5 ( right ) . broadly speaking ,
the simulations evolve similarly to the vpc models ; there are two cavities forming in opposite directions and expanding , with material falling in along the equatorial plane .
overall , the differences are much smaller than between the tk and vp models presented above .
this happens because with the non - spherical injection of feedback , a channel appears along the equatorial plane which allows inflow and outflow to coexist
. therefore , dense gas no longer accumulates at the edge of a well - defined outflow bubble , but is brought to the midplane by the same processes that act in the vpc models .
a similar structure of dense inflow along the midplane and diffuse outflow in the polar direction develops , with some irregularities due to the turbulent gas density distribution . in the l1 and l2 simulations , both tkc and vpc models produce similarly - sized bubbles , while the brightest agn produces a significantly larger bubble in the vpc simulation .
this last size difference is exacerbated by the fact that the bubble reaches the edge of the initial gas distribution in the vpc simulation , and its expansion accelerates . at earlier times , the difference in size between tkc - l5 and vpc - l5 bubbles is much less pronounced . despite the morphological differences and similarities , we are more interested in the effect that agn feedback would have upon the evolution of the host galaxy . therefore
, we now consider several integrated parameters of the system and their time evolution .
one of the most important effects that agn feedback has upon its host is the energy injection , which leads to gas heating and expulsion .
the heated gas then cools down and forms structures including clumps and filaments , as seen in the morphology plots .
in order to better understand the difference between thermal and virtual particle feedback prescriptions , we show the phase diagrams of the four l2 simulations in figure [ fig : phase_l2 ] and correspondingly for the l5 simulations in figure [ fig : phase_l5 ] . in these plots , colour represents the density of particles in the particular parts of the plot , with blue being lowest and red being highest .
dashed lines indicate the bremsstrahlung cooling times
. a clear difference can be seen between the tk - l2 and vp - l2 models ( figure [ fig : phase_l2 ] , top row ) .
thermal feedback creates slightly more dense gas , which cools down quickly to the two horizontal branches of the equilibrium temperature curve , between @xmath22 and @xmath41 k. meanwhile , virtual particle feedback acts upon diffuse gas and heats it up to @xmath6-@xmath42 k. this gas has long cooling times , @xmath43 yr and therefore stays hot and flows outward .
thus the virtual particle feedback prescription allows the gas to retain a larger fraction of the agn input energy than the thermal prescription .
such difference is much less pronounced in the conical feedback runs ( bottom row ) , leading to much more similar morphology .
the difference between the tk - l5 and vp - l5 models is even more pronounced ( figure [ fig : phase_l5 ] , top row ) , as the thermal feedback compresses gas close to the smbh into a very dense shell ( @xmath44 k , @xmath45 g @xmath46 ) .
meanwhile , the virtual particle feedback creates a large population of gas with @xmath47 k , which expands rapidly outward .
once again , the conical feedback simulations ( bottom row ) are much more similar to each other , with large amounts of hot expanding gas and much less extremely dense material . these differences in energy retention are clearly visible when the total gas energy is considered . in figure
[ fig : energy_evolution ] , we show the change of total particle energy compared with input energy as a function of time : @xmath48 here , the gas energy @xmath49 includes gravitational potential , thermal and kinetic energy .
we do not plot the results of the l1 simulations , because they are qualitatively very similar to those of l2 , shown in the top panel .
the l5 models are in the bottom panel .
virtual particle simulation results are shown in solid lines ( black for spherical , blue for conical virtual particle emission ) , while the thermal feedback simulations are shown in dashed lines ( red for spherical , green for conical feedback injection ) . in both cases ,
the conical feedback simulations lose less energy to cooling than spherically symmetric ones .
this happens because with conical feedback , dense gas is pushed toward the midplane and the heated gas is mostly diffuse , therefore has a low cooling rate ( the presence of hot diffuse gas is also seen in the phase plots above ) . in the l2 models ,
this low cooling rate is still high enough to make the gas lose most of the injected energy , leaving only a few percent as the internal energy .
however , this is still a factor few better retention than the spherically symmetric model in the virtual particle case ; the spherically symmetric thermal feedback model has gas energy decreasing to values even lower than those at @xmath50 myr , as gas falls deeper into the potential well .
it is worth noting that in the tkc - l2 and vpc - l2 simulations , gas retains very similar amounts of input energy . in the l5 simulations ,
the virtual particle conical feedback model is essentially purely adiabatic , losing only @xmath51 of the input energy by @xmath52 myr ; the thermal conical feedback model loses a much larger fraction of its energy , since it does not create such a large bubble .
even so , the difference between spherical and conical feedback models is significant : energy retention improves by a factor 5 or more in the thermal model .
even though the high - luminosity thermal feedback simulation retains significantly less energy than its virtual particle counterpart , the similarity in the l2 models is encouraging .
below we show that similar trends exist in other integrated quantities .
another particularly important aspect of the co - evolution of smbhs and their galaxies is the agn duty cycle , i.e. the frequency and duration of agn activity episodes . in a realistic system
, one might expect agn feeding and feedback to occur simultaneously , at least up to some critical agn luminosity . within the agn wind feedback model
, this happens because most of the agn wind energy is carried away by low - density gas , allowing dense clouds to fall in toward the smbh .
although our simulations do not resolve the scales of smbh feeding , we nevertheless can investigate the ability of various feedback prescriptions to reproduce simultaneous inflows and outflows .
for this , we plot the radial profiles of gas inflow and outflow rates in the l2 and l5 simulations in figures [ fig : massflow_profiles_l2 ] and [ fig : massflow_profiles_l5 ] , respectively .
the plots are made at three times for each simulation , @xmath53 myr , @xmath37 myr and @xmath54 myr after the agn switches on .
the inflow rate is defined as @xmath55 where @xmath56 kpc is the thickness of the radial bin and the sum goes over all particles with radial velocity directed inward and higher in magnitude than half the background velocity dispersion .
the second equality shows that the inflow rate can be expressed via the mean inflowing gas density @xmath57 and does not depend on the choice of @xmath58 so long as the bin is thick enough to sample a large number of particles .
we neglect particles with small negative velocities because those are dominated by turbulent motions rather than inflow or outflow .
the outflow rate is defined in the same way , but the sum is made over all particles with radial velocities greater than half @xmath59 .
the four l2 models all generally have stronger inflows than outflows .
this is expected , since the agn luminosity is not large enough to drive away all the gas by momentum push alone .
however , clear differences are visible among the models , with tk - l2 having essentially zero outflow , vp - l2 having outflow rates @xmath60 times lower than inflow , tkc - l2 having outflow rates only @xmath61 times smaller , especially in the region @xmath62 kpc , and finally vpc - l2 having inflow and outflow rates of comparable magnitude .
we see that the qualitative difference present between the tk - l2 and vp - l2 models disappears when feedback is injected conically , even though a notable quantitative difference remains .
it is also worth noting that the morphology of all four l2 models appears very similar in the outer regions , since the outflows are never fast enough to extend beyond @xmath63 kpc in 1 myr .
the l5 models display much greater differences among themselves .
the tk - l5 simulation produces very little outflow and a strong inflow , rather similar to its lower - luminosity counterpart .
however , the inflow proceeds only to @xmath64 kpc and stalls there . in the other three simulations , outflows dominate , initially in the inner regions ( black lines ) , but later throughout the simulation volume .
the two conical feedback simulations have qualitatively very similar radial profiles of both inflow and outflow , echoing the similarity in their morphologies ( figures [ fig : vpc_morph ] and [ fig : tkc_morph ] ) .
it is important to note that even though the outflow is very powerful , the inflow rate in the central @xmath65 kpc is similar to the outflow rate , showing the the smbh may be fed even while producing such a massive outflow .
we also plot the time evolution of the rate of gas particle accretion by the smbh particle ( figure [ fig : mdot_bh ] ) and the total rate of gas outflow , approximated as @xmath66 ( figure [ fig : mdot_out ] ) .
the line colours are as in figure [ fig : energy_evolution ] . in all models ,
there is no accretion for the first @xmath67 myr of agn activity , because the gas is falling toward the significantly reduced accretion radius of the smbh particle . here
, the two spherically symmetric models are more similar than when the energy retention is considered . however , there is an important qualitative difference . in the tk - l5 model ,
thermal feedback is able to shut off accretion entirely , by keeping the gas at the edge of a well - defined bubble ( see figure [ fig : tk_morph ] , right panel ) . on the other hand , the virtual particle model , while producing very large outflow bubbles , also allows some gas to fall on to the smbh particle .
the accretion rate is comparatively small , hardly exceeding @xmath68 yr@xmath69 , but the possibility of accretion is significant . in the l2 models ,
the situation is reversed : thermal feedback suppresses accretion for @xmath70 myr , but then the outflow bubble collapses and accretion rate rises rapidly to @xmath71 yr@xmath69 .
meanwhile , the vp - l2 models has a steadily rising accretion rate , without such sudden changes .
both high - luminosity conical feedback models , unsurprisingly , allow far more accretion on to the smbh .
simulation vpc - l5 is more efficient at pushing gas away than the tkc - l5 simulation , therefore the accretion rate is significantly lower , only rising to a peak of @xmath72 yr@xmath69 .
the thermal feedback model , on the other hand , allows accretion at rates approaching @xmath73 yr@xmath69 at @xmath74 myr .
in the low - luminosity models , conical feedback actually decreases the gas inflow rate by a factor few .
this happens because in these models , outflow bubbles form and prevent accretion in some directions , channeling accreting material toward the midplane .
the tkc - l2 simulation has a higher accretion rate at first , since it is initially more efficient in compressing the gas toward the midplane , but from @xmath75 myr onward , the two accretion rates become very similar .
the difference among the gas outflow rates ( figure [ fig : mdot_out ] ) is far more striking .
the spherical thermal feedback model is incapable of driving a significant outflow in either low- or high - luminosity models , as most of the agn input energy is radiated away , and the gas is kept in a dense , but small and hardly expanding , shell .
the spherical virtual particle simulation produces an outflow with at least several times higher mass flow rate , and in the higher luminosity case , the outflow rate approaches that of the conical simulations by @xmath76 myr .
this happens because of the complex morphology present in the virtual particle simulation , with outflowing gas being diffuse , and hence reaching very high velocities .
the conical feedback models produce significant outflows , essentially because the dense gas is channeled away and diffuse gas can be accelerated to high velocities .
the mass outflow rates approach @xmath77 yr@xmath69 in the l5 models .
it is interesting that the collimation of feedback into a cone actually increases the outflow rate in the virtual particle simulations , most likely because conical feedback produces more spatially distinct regions of inflow and outflow , so there is less mixing between inflowing and outflowing gas , which would lead to lower outflow rates .
the thermal feedback model produces a significantly weaker outflow than the virtual particle model in the conical case as well , but the ratio between outflow rates is much smaller than for the spherical feedback injection simulations . in figure [
fig : rho_dist ] , we plot information regarding the evolution of gas density distribution during the agn activity episode .
more precisely , we plot the values of the 10th , 25th , 75th and 90th percentile of the gas density distribution , in order to evaluate how the different phases of gas ( diffuse , medium and dense ) behave during the agn activity episode .
the line colours and styles are the same as in previous time evolution figures .
we see that the diffuse gas evolves almost identically for the whole agn activity episode in the l2 models and for the first @xmath78 myr in the l5 models , mostly because most of this gas is far away from the smbh and is not affected by the agn outflow at first .
none of the l2 models produces a bubble extending to the outskirts of the gas sphere , so there is little evolution of the diffuse gas in this model throughout the time period considered ; the same is true for the tk - l5 model . in the other three models ,
diffuse gas starts to become even more diffuse as the outflow bubble moves past it and disperses the gas in a larger volume .
the dense gas becomes progressively more dense , but with certain differences among the simulations .
the virtual particle models compress the dense gas by @xmath79 times in the period between @xmath80 myr , corresponding to the formation of a complex inflow - outflow structure with many dense clumps and filaments .
the tkc - l5 model shows a similar behaviour for the densest gas , although the compression takes a slightly longer time .
once this structure is in place , however , density does not increase further , as the densest material either turns into sink particles or is accreted on to the smbh .
the tk - l5 model , on the other hand , shows essentially unrestricted density growth , because here , the densest gas is also the warmest , because it receives all the agn energy input , and therefore can not fragment .
the tk - l2 model behaves in the same fashion initially , but later the dense gas collapses on to the smbh and is accreted , resulting in a drop in the densest gas density values .
our suite of 13 simulations allowed us to explore the effect that the subgrid agn feedback prescription has upon the evolution of a turbulent gas sphere .
our main results are as follows : * spherically symmetric thermal agn feedback is unable to reproduce the complex structures found in virtual particle models .
the reason is that the thermal feedback prescription injects energy only into the surrounding gas , creating a well - defined bubble around the smbh . *
the virtual particle prescription allows for simultaneous gas inflow and outflow , as well as formation of dense gas filaments and clumps , which are not found in the thermal feedback case . * replacing both prescriptions with biconical versions results in a marked improvement in their agreement .
in particular , the morphologies of the two models then become very similar , with clear outflow bubbles coexisting with inflow along the equatorial plane . * the conical feedback models are qualitatively similar to the corresponding spherical virtual particle simulations , but quantitatively very different from both spherical models : the outflow bubbles produced by conical feedback are generally larger , the outflow rates higher and much more of the agn input energy is retained by the gas . *
the fraction of agn input energy retained in the gas is very low in the spherically symmetric thermal feedback simulation , even at high agn luminosity , but increases to @xmath81 in the conical high - luminosity one . in the virtual particle models ,
the energy retention fraction increases to an ever higher value of @xmath82 . *
spherically symmetric thermal feedback either does not allow any accretion on to the smbh , or allows most of the gas to fall in .
it also produces only a weak outflow .
virtual particle and conical models produce both significant inflow and outflow at the same time .
although the set up of our simulations is strongly idealised , the results have several implications for the understanding of agn activity and its effects upon the host galaxy .
we discuss those implications below .
as mentioned in the introduction and model sections , the commonly - used thermal - kinetic agn feedback models suffer from a number of drawbacks .
the most important drawback is the assumed spherical symmetry , i.e. energy injection into gas close to the smbh .
if it happens that there is gas only to one side of the smbh , all the energy will be injected into this gas , rather than spread around symmetrically .
this creates a paradoxical situation where assumed spherical symmetry results in energy injection which is often very far from symmetric .
the problem with this energy injection in generating outflows lies in the fact that once any parcel of gas escapes far from the smbh , it is no longer affected by feedback .
therefore , as far as most gas particles in the simulation are concerned , feedback is very intermittent , and this leads to much weaker outflows than would be possible if feedback acted upon a gas parcel for a prolonged period of time .
a small amount of gas does experience this prolonged feedback , however , it gets squashed between agn feedback on one side and inflowing gas on the other . gas density therefore increases , cooling becomes more efficient and progressively more of the feedback energy is lost to radiation rather than used to drive the outflow .
therefore , the resulting agn outflow is much weaker than it would otherwise and more energy is required to remove the gas from the host galaxy than it would be if these multi - phase effects were treated properly .
these issues do not necessarily impact the ability of cosmological simulations to reproduce the mass functions and growth histories of present - day galaxies and smbhs , even though the latter requires appropriate tuning of feedback parameters @xcite .
agn accretion and feedback self - regulate in these simulations , so that a less efficient agn feedback allows more accretion which in turn increases the outflow rate , having a negligible effect upon the long - term evolution of the total stellar mass in the galaxy @xcite .
however , as such models , as well as models of isolated galaxies , become more detailed , they start reproducing other , smaller - scale , processes and structures , such as the phase structure of the gas , the spatially - resolved history of star formation , the duration of agn episodes , and so on . all of these , and many other , properties of the galaxy are affected by the details of how agn feedback interacts with the ism , and therefore understanding the limitations imposed by the numerical feedback scheme is of paramount importance when interpreting the results of these models .
ideally , one would like to minimize the limitations of the numerical method .
our proposed scheme of injecting feedback energy into a cone alleviates the two biggest issues of thermal feedback prescriptions described above .
while it does not really resolve the interaction of the agn outflow with the multiphase gas , it creates a possibility for the gas to remain in two phases ( cold dense inflowing and hot diffuse outflowing ) while interacting with the outflow .
dense gas is quickly channeled to the bottom of the effective potential well , i.e. to the regions where feedback does nt counteract gravity . once there
, dense gas can form filaments which feed the smbh or fragment into self - gravitating clumps . meanwhile ,
diffuse gas gets heated and pushed by the feedback energy injection and forms a massive outflow which cools slowly due to low gas density .
this results in an adiabatic outflow , consistent with the analytical predictions of the wind feedback model . given that a conical thermal feedback prescription is capable of resolving a lot of the complexity of simultaneous inflows and outflows with little extra computational cost
, we suggest that it may be interesting to explore its effects in cosmological simulations , which self - consistently track the growth and energy release from agn over long timescales
. this would allow one to constrain the agn growth scenarios , the @xmath23 relation , the effects of agn outflows upon galaxies and clusters , the triggering of star formation , as well as the history of smbh growth , with better precision than is currently available @xcite .
one particularly interesting consequence might be that the presence of simultaneous inflows and outflows around smbh change the qualitative behaviour of agn self - regulation ( see section [ sec : selfreg ] below ) .
a few important aspects have to be considered when implementing this scheme into cosmological simulations .
the first is numerical resolution , which is typically much worse in cosmological simulations than it is in the simulations presented here . in a previous paper @xcite
, we investigated the importance of resolution upon the efficiency of feedback and found that low - resolution simulations produce feedback that is much more negative .
this happens in part because those simulations are unable to resolve the complex structures developing in the turbulent ism . with a conical feedback prescription
, it should be easier to resolve those structures , since the cavities and filaments are larger ( compare the right panels in figures [ fig : vp_morph ] and [ fig : vpc_morph ] ) .
however , we may still expect a weaker overall effect of agn feedback due to the fact that the gas density contrast is not resolved as well , leading to overcooling of the hot gas in the bubbles .
another aspect is defining the appropriate direction of the cone s axis .
we suggest that this axis should coincide with the axis of rotation of the smbh particle .
rotation can be tracked by tracking the angular momentum magnitude and direction of the infalling gas , and the plane of the accretion disc should be perpendicular to this axis of rotation . since the agn outflow originates in the disc , it is reasonable to assume that it goes in the polar direction @xcite .
the opening angle of the wind cone is a free parameter , which might be constrained by more detailed models , or investigated empirically .
however , such an investigation is beyond the scope of this paper .
finally , cosmological simulations contain galaxies which have varied structures , including discs , large clumps and cavities created by previous outflows .
the interaction of a new outflow with these structures may be much more complex than in our simulations presented above .
for example , one directed conical outflow might intercept a dense clump and lose a large fraction of its energy , while another may expand into a pre - existing cavity , having little effect upon the mass budget in its host galaxy .
a more thorough investigation of feedback prescriptions in a more realistic galaxy model should be done before implementing them into a cosmological simulation .
we intend to perform such tests in a future publication .
our simulations use a fixed agn luminosity rather than keeping it tied to the central particle accretion rate .
given that the central accretion rates in different simulations can differ by orders of magnitude ( see figure [ fig : mdot_bh ] ) , this may seem as a significant drawback of the numerical setup . however , we feel that simulations addressing just the feedback aspect of agn activity are still very useful since the models with feedback determined by bh accretion rate employ a number of assumptions about this rate . given the non - linear accretion - feedback connection , it would be hard to arrive at definite conclusions about the feedback efficiency in these more complex and less constrained simulations .
in addition , the actual expected evolution of the system is somewhat more complicated than the accretion rate of the central particle would suggest .
infalling gas contains some angular momentum , and therefore forms an accretion disc around the smbh , rather than falling in directly .
given that in the simulations the accretion rate is measured at a radius of 10 pc , most of this accreting gas will take far longer than @xmath16 myr to reach the smbh , by which time the activity episode in our test run is over .
additionally , some of this gas will fragment and form a nuclear star cluster rather than feeding the smbh @xcite , further reducing the impact upon the agn luminosity .
therefore , our simulations may be seen as an investigation of how a large - scale gas distribution is affected by agn activity caused by accreting a cloud of predetermined mass , from @xmath83 ( l1 simulations ) to @xmath84 ( l5 simulations ) . in a more realistic simulation ,
long - term self - regulation of agn activity becomes important .
here , the temporal evolution of simulations using the thermal - kinetic and virtual particle feedback prescriptions might be very different from one another . when the agn luminosity increases , simulations using the spherically - symmetric thermal - kinetic agn feedback prescription abruptly change from a situation where no outflow is occurring to one where the outflow completely shuts off agn accretion .
this latter situation may reverse after some time , as the material piles up at the edge of the outflow bubble , and its weight may overcome the force produced by agn feedback , leading to collapse and resuming of accretion .
however , while the bubble holds , there is no accretion .
conversely , while there is accretion , no outflow bubble can exist .
this `` all - or - nothing '' situation places tight constraints on the regulation of agn accretion rate . when the agn is fed at a large enough rate , an outflow forms and shuts off accretion .
as the accretion rate drops to zero ( with perhaps some delay imposed in the simulation to mimic the draining of the sub - resolution accretion disc ) , agn luminosity also decreases and feedback stops , allowing accretion to resume .
we may therefore expect agn activity to keep switching on and off , with the duration of the `` on '' phase comparable to the viscous timescale of the accretion disc ( @xmath85 yr ) and the `` off '' phase lasting for as long as it takes for gas to fall back to the smbh ( @xmath86 yr ) .
these timescales are comparable to some estimates of the lifetimes of agn phase @xcite , but this is only coincidental , since the processes described here happen due to numerical , rather than physical , reasons .
the virtual particle prescription , as well as both conical prescriptions explored in this paper , allow simultaneous inflow and outflow to coexist . at some point ,
the larger gas reservoir gets depleted and accretion switches off , leading to a longer period of inactivity than in the case of spherical thermal feedback .
this simulation therefore does not manifest an artificial short - timescale `` flickering '' and can be used to investigate the physical reasons of the switching on and off of agn activity . on the other hand ,
some integrated parameters , such as the total mass accreted by the smbh or expelled from the galaxy , may not differ very much between thermal - kinetic and virtual particle feedback simulations .
the reason for this is simply that on long timescales smbh accretion tends to self - regulate so that the total energy release into the surrounding medium is just enough to balance gravity @xcite .
a thermal - kinetic feedback prescription would lead to faster accretion followed by a stronger burst of agn activity , which leads to a larger feedback bubble and shutting off of accretion for a longer time .
the virtual particle prescription would result in a more uniform accretion and outflow rate , with neither undergoing such strong changes as in the thermal - kinetic feedback model . as a result
, eventually the results of the two simulations might look more similar to each other than might seem from our simulation results .
still , on short timescales , the morphology and temporal variability of inflow and outflow rates would differ significantly among the simulations .
the possibility of having prolonged periods of agn activity coincident with smbh growth via accretion also has an implication for the growth of the very first smbhs .
these black holes are known to have masses exceeding @xmath87 by @xmath88 , when the universe was @xmath89 yr old ( e.g. * ? ? ?
if these smbhs grew from stellar - mass progenitors , their growth rate must have been close to the eddington limit for most of their lives @xcite .
the possibility of having rapid gas inflows during the agn phase suggests that this might be easier than previously thought and eliminates one of the potential drawbacks of this growth mechanism in explaining the large observed masses .
we note , however , that it is possible to grow black holes to the required masses at high - z even with the typical thermal - kinetic agn feedback prescription @xcite , at least for seed smbh masses of @xmath90 .
therefore a change in our understanding of accretion and feedback is not necessary to explain the high - z quasars .
thermal feedback simulations , both spherically symmetric and conical , eliminate some of the structure present in turbulent gas .
meanwhile , virtual particle simulations , and to some extent the conical thermal feedback simulations , enhance some of the structures , by enveloping dense gas in hot diffuse bubbles and compressing it more than self - gravity alone would .
this leads to formation of more clumps and more efficient star formation within them @xcite , resulting in a higher rate of fragmentation and star formation .
in other words , agn activity has a positive feedback upon star formation in the host galaxy .
this effect has been explored in previous numerical works .
@xcite showed that the outflowing material can fragment as it cools down , leading to potentially rapid star formation and stars ejected from host galaxies .
@xcite explored the interaction of an agn jet with a clumpy ism and found that the jet compresses dense gas , prevents adiabatic re - expansion of the clumps , and enhances the sfr of the whole galaxy .
there is also tentative observational support that galaxies with more luminous agn have higher star formation rates and efficiencies ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , although the causal connection is not certain .
we chose not to consider in detail the fragmentation of gas in our simulations , as that is not the main point of the present study .
nevertheless , we find qualitatively similar effects , but , importantly , they are much more significant in the virtual particle feedback models , where the multiphase ism is affected by feedback based on its density . in a more realistic simulation ,
the effects of agn outflow , and of the numerical prescription used to track its interaction with the ism , might significantly affect the fragmentation rate and spatial locations in the galaxy .
we suggest that using a more detailed agn feedback prescription , such as the conical thermal feedback one , would allow investigating this positive feedback in a far more realistic way . one implicit assumption we made so far when analysing the results is that the virtual particle feedback prescription is a good representation of reality .
the method certainly has several advantages over the thermal feedback model .
it is truly isotropic , i.e. feedback energy is emitted in all directions from the agn and interacts with the ism in all directions independently of the shape of the cavity surrounding the agn .
in addition , the model , by construction , accounts for the different optical depths of gas in different directions , leading to situations where a dense clump can shield the gas behind itself from being blown away .
on the other hand , the model suffers from some shortcomings as well .
the interaction cross - section between virtual particles and sph particles is a free parameter of the model .
this means that the depth to which virtual particles penetrate the gas can vary significantly between simulations . as a result
, dense gas clumps can sometimes be obliterated unphysically rapidly .
conversely , in other cases , gas might be compressed too strongly where a gentler push would be more realistic and represent the many filaments forming along the contact discontinuity between the agn wind and the shocked ism .
another drawback is that the virtual particles are assumed to always propagate radially out from the agn .
this is unrealistic in cases of uneven density , where a steep density gradient may be present in a direction not parallel to the direction of propagation of the virtual particle . in this case
, a realistic interaction would have the virtual particle flying off at an angle to its initial direction , something which the present model does not allow .
this drawback may have many subtle effects in how the gas morphology is affected by agn activity .
given these drawbacks , we stress that the results of the virtual particle simulations should not be considered as perfect representations of reality .
however , given that they reproduce a far more complex morphology of the turbulent ism and properties of outflows in agreement with observations , we believe that their results represent real galaxies better than those of the spherical thermal feedback models .
in this paper , we presented a number of simulations of agn feedback affecting a turbulent gas sphere .
the simulations are designed to show the impact of different subgrid recipes of agn feedback upon the evolution of the whole system .
we show that the commonly - used spherically - symmetric thermal agn feedback prescription is unable to reproduce the variety of complex structures observed in more detailed models , fails to reproduce the observed rapid massive outflows and in general has a much weaker effect on the host galaxy than predicted by analytical calculations of feedback models .
on the other hand , a relatively simple improvement to the prescription - injecting energy into a cone rather than a sphere - alleviates most of these issues .
the outflows become rapid , massive and close to adiabatic , dense structures and inflowing filaments coexist with outflowing gas , and gas fragmentation rates are significantly increased . in particular ,
even extremely bright agn are unable to shut off gas accretion completely , allowing the smbh to continue growing .
we suggest that this improvement could be used in cosmological simulations in order to better model the effects that agn have on their host galaxies and clusters .
kz is funded by the research council of lithuania grant no . mip-062/2013 .
mab and sn acknowledge an stfc grant .
mab is funded by a stfc research studentship .
we thank justin read for the use of sphs .
this research used the dirac complexity system , operated by the university of leicester it services , which forms part of the stfc dirac hpc facility ( www.dirac.ac.uk ) .
this equipment is funded by bis national e - infrastructure capital grant st / k000373/1 and stfc dirac operations grant st / k0003259/1 .
dirac is part of the uk national e - infrastructure .
r. a. , schaye j. , bower r. g. , furlong m. , schaller m. , theuns t. , dalla vecchia c. , frenk c. s. , mccarthy i. g. , helly j. c. , jenkins a. , rosas - guevara y. m. , white s. d. m. , trayford j. w. , 2015 , mnras , 450 , 1937 | agn feedback is an important ingredient in galaxy evolution , however its treatment in numerical simulations is necessarily approximate , requiring subgrid prescriptions due to the dynamical range involved in the calculations .
we present a suite of sph simulations designed to showcase the importance of the choice of a particular subgrid prescription for agn feedback .
we concentrate on two approaches to treating wide - angle agn outflows : thermal feedback , where thermal and kinetic energy is injected into the gas surrounding the smbh particle , and virtual particle feedback , where energy is carried by tracer particles radially away from the agn .
we show that the latter model produces a far more complex structure around the smbh , which we argue is a more physically correct outcome .
we suggest a simple improvement to the thermal feedback model - injecting the energy into a cone , rather than spherically symmetrically - and show that this markedly improves the agreement between the two prescriptions , without requiring any noticeable increase in the computational cost of the simulation .
quasars : general accretion , accretion discs ism : evolution methods : numerical |
the kohn - sham ( ks ) density functional theory ( dft ) @xcite is one of the most widely used electronic structure theories for atoms , molecules and solids in various areas of physics , chemistry and molecular biology due to its computational efficiency and useful accuracy .
it simplifies a many - electron wave - function problem to an auxiliary one - electron problem , with only its exchange - correlation part carrying the many - electron effects to be approximated in practice . among numerous exchange - correlation approximations , the local spin density approximation ( lsda ) @xcite ,
the standard perdew - burke - ernzerhof ( pbe ) generalized gradient approximation ( gga ) @xcite , and the becke-3-lee - yang - parr ( b3lyp ) @xcite hybrid gga dominate the user market of dft @xcite .
the former two are efficient semilocal functionals widely used for extended systems , while the latter is a computationally more expensive nonlocal functional that hybridizes a gga with the exact exchange energy and is popular for finite systems . at the semilocal level , however , meta - gga ( mgga ) is the highest rung of the so - called jacob s ladder in dft @xcite and potentially the most accurate one @xcite , which can also serve as a better base for hybridizing with the exact exchange energy .
semilocal approximations ( e.g. , refs .
@xcite ) of the form e_xc^sl[n _ , n_]=d^3r n _ xc^sl(n_,n _ , n_,n _ , _ , _ ) [ semi_local ] require only a single integral over real space and so are practical even for large molecules or unit cells . in eq . , @xmath0 and @xmath1 are the electron densities of spin @xmath2 and @xmath3 , respectively , @xmath4 the local gradients of the spin densities , @xmath5 the kinetic energy densities of the occupied ks orbitals @xmath6 of spin @xmath7 ( @xmath8 ) , and @xmath9 the approximate exchange - correlation energy per electron .
all equations are in atomic units . in addition to @xmath0 and @xmath1 , the only ingredients in lsda , ggas also use the density gradients ; mggas additionally include the kinetic energy densities @xmath10 . with the encoded information of shell structures from the kinetic energy densities
, mggas can distinguish different orbital - overlap regions @xcite , and thus deliver simultaneous accurate ground - state properties for molecules , surfaces , and solids which could nt be obtained with a gga or lsda @xcite .
computationally , mggas are not much more expensive than lsda or gga @xcite . in computations for molecules containing transition - metal atoms ,
the tao - perdew - staroverov - scuseria ( tpss ) @xcite mgga is only @xmath11 slower @xcite than pbe . to understand the performance of mggas , a recent study @xcite investigated the effect of @xmath12-dependence on mggas through the inhomogeneity parameter @xmath13
here , @xmath14 , @xmath15 is the von weizs@xmath16cker kinetic energy density , and @xmath17 is the kinetic energy density of the uniform electron gas ( ueg ) .
@xmath18 is a lower bound on @xmath12 with @xmath19 only for a single - orbital region @xcite .
therefore , @xmath20 measures locally how far @xmath21 deviates from being of single - orbital character on the scale of the ueg , and thus characterizes the extent of orbital overlap . besides the ability of @xmath22 to distinguish different orbital - overlap regions , ref
. further showed that the effect of the @xmath22-dependence on mggas is qualitatively equivalent to that of the dependence on the reduced density gradient @xmath23 $ ] , another dimensionless inhomogeneity parameter that measures how fast and how much the density varies on the scale of the local fermi wavelength @xmath24 with @xmath25 .
the @xmath26-dependence is well understood at the gga level @xcite , which mggas inherit @xcite . as a result of the study on the @xmath12-dependence @xcite , a simple exchange functional , where the @xmath22-dependence is disentangled from the @xmath26-dependence ,
was constructed as an interpolation between the single - orbital regime ( @xmath27 ) and the slowly - varying density regime ( @xmath28 ) @xcite , which we will discuss more in section [ sect : construction ] . in analogy to the monotonically increasing @xmath29dependence that ggas are often designed to have to favor more inhomogeneous electron densities , the simple exchange functional has a monotonically decreasing @xmath30dependence to favor more the single - orbital regions . when combined with a variant of the pbe correlation @xcite , the resulting mggas ( denoted as mgga@xmath31ms with ms standing for `` made simple '' ) @xcite performs equally well for atoms , molecules , surfaces , and solids , with an overall performance that is comparable to the sophisticated revised tpss ( revtpss ) mgga @xcite
moreover , mgga_ms yields excellent binding energies for the w6 water clusters @xcite , even better than those from the highly parametrized m06-l mgga @xcite whose training sets include noncovalent interactions ( hydrogen bonds and van der waals interactions ) . further tests on general main group thermochemistry , kinetic , and noncovalent interactions @xcite showed that mgga_ms , though not as good as m06-l , systematically improves the descriptions for the noncovalent interactions over conventional semilocal functionals , e.g. , pbe , tpss , and revtpss . by including training sets of noncovalent interactions ,
the m06-l mgga was trained to capture medium - range exchange and correlation energies that dominate equilibrium structures of noncovalent complexes @xcite .
for example , m06-l delivers good binding energies for the s22 set that is dominated by noncovalent interactions and will be discussed more in sec .
[ sect : results ] , good structural and energetical properties of layered materials bound by van der waals interactions ( e.g. , graphite , hexagonal boron nitride , and molybdenum disulfite @xcite ) , and good adsorption energies of aromatic molecules adsorbed on coinic metal surfaces @xcite .
m06-l represents one of the developments of mggas that are fitted to a variety of training sets with numerous parameters to try to improve different properties together @xcite .
however , it should be stressed that too many fitting parameters could cause problems , e.g. , convergence of calculations and the related smoothness of binding curves @xcite , and the prediction for systems far from training sets might not be realistic @xcite .
semilocal ( sl ) approximations can be reasonably accurate for the near - equilibrium and compressed ground - state properties of `` ordinary '' matter , where neither strong correlation nor long - range van der waals interaction is important .
however , stretched bonds that arise in transition states of chemical reactions or in the dissociation limits of some radical or heteronuclear molecules , etc .
, require full nonlocality @xcite .
the global hybrid functionals discussed later can provide such needed nonlocality .
long - range van der waals interactions , originated from spontaneous correlations between two distant electron densities , also require full nonlocality @xcite , which is different from the previous one @xcite . to obtain long - range van der waals interactions , there are various dft - based dispersion techniques , including the post - dft empirical pairwise potential corrections @xcite and the van der waals density functionals ( vdw - dfs ) @xcite , which are based on the pairwise additivity and problematic for metallic systems @xcite .
see ref . for an overview of the progess on dft - based dispersion techniques .
conventional semilocal functionals , e.g. , the widely used pbe gga , even lack the ability to describe noncovalent interactions near equilibrium , while the m06-l mgga and mgga_ms as well as its variant proposed here improve over pbe .
the global hybrid ( gh ) idea , due to becke @xcite , introduces some full nonlocality into the calculation but only at the level of @xmath32 , which can be evaluated semianalytically from the kohn - sham orbitals in some computer codes . in its simplest ( one parameter ) version @xcite : e_xc^gh = ae_x^exact+(1-a)e_x^sl+e_c^sl [ global_hybrid ] where the exact - exchange mixing parameter @xmath33 takes an empirical value in the range @xmath34 . a rough but convincing argument @xcite for why eq .
( [ global_hybrid ] ) works is : semilocal exchange - correlation typically overestimates atomization energies and underestimates energy barriers of chemical reactions , while exact exchange without correlation makes errors of the opposite sign , so that mixing of the two will yield better atomization energies and barriers than either alone .
therefore , values of @xmath33 vary with the choice of semilocal functionals when fitting eq .
( [ global_hybrid ] ) to formation enthalpies .
the more a semilocal functional overbinds molecules , the larger the value of @xmath33 is .
et al . _
@xcite found @xmath33 to be 0.60 for the pbesol gga , 0.32 for the pbe gga , and 0.1 for the tpss and revtpss mggas .
improving the underlying semilocal functional reduces the value of @xmath33 needed to fit the atomization energies or enthalpies of formation , which however can worsen the barrier heights . some global hybrids
@xcite thus employ highly fitted semilocal parts intended not to be accurate by themselves but to work well with a fraction of exact exchange .
the popular b3lyp @xcite hybrid gga has a somewhat more complicated form and reads : @xmath35 where @xmath36 @xcite is the lsda correlation energy in the random phase approximation which does not yield the correct uniform electron gas limit . in the equation , a=0.2 , b=0.72 , and c=0.81 . if we use the similar idea as in eq . to define the underlying semilocal functional for b3lyp by replacing @xmath32 with @xmath37
, we end up with a gga , denoted as blyp@xmath38 , @xmath39 .
blyp@xmath38 differs from blyp in reducing the gradient correction of the b88 exchange to ( a+b)=92% and mixing the vwn3 and lyp correlations .
this difference confirms the rough argument of the previous paragraph . typically , blyp underbinds molecules , and inclusion of exact exchange worsens the underbinding .
the reduction of the gradient correction of the b88 exchange and the mixture of the vwn3 and lyp correlations then turn blyp@xmath38 into an overbinding functional suitable for hybridization with the exact exchange .
the idea of tuning the gradient correction of b3lyp can be transfered to the construction of our hybrid mgga , where both the @xmath29dependence and the @xmath30dependence are tuned . in the next section
, we will present the computational details , followed by the constructions of semilocal and hybrid mggas in section [ sect : construction ] .
results and discussions will be given in section [ sect : results ] .
and we give the conclusions in section [ sect : conclusions ] .
the functionals described here were implemented in the development version of the electronic structure program @xcite .
all calculations employ the fully uncontracted 6 - 311++g(@xmath40 ) basis set @xcite to obtain benchmark quality results .
we used the ultrafine grid with 99 radial shells and 590 angular points for numerical integration of the dft xc potential .
the program was used for all calculations presented in this study .
different training sets were used to adjust empirical parameters . for the ae6 and bh6 tests
sets @xcite we employed reference data from highly accurate quantum - chemical calculations @xcite .
experimental reference values were used for g2/97 ( 148 molecules ) @xcite and bh42/03 ( 21 forward and reverse hydrogen transfer barrier heights ) @xcite .
the fitted parameters would not change much if the highly accurate quantum - chemical reference data from ref . would have been used .
the superset of g2/97 and g3 - 3 ( 75 molecules ) @xcite ( a total of 223 molecules comprising the g3/99 test set ) @xcite , the s22 test set for weak interactions @xcite , as well as the barrier height test sets htbh38/04 ( 19 forward and reverse hydrogen transfer barrier heights ) @xcite and nhtbh38/04 ( 19 forward and reverse non - hydrogen transfer barrier heights ) @xcite are employed for assessment of the fitted functionals .
the employed geometries and reference values are available from the supporting information of refs . and .
the counterpoise correction @xcite was used to reduce the basis set superposition errors for calculations of weak interactions . a vibrational scale factor of 0.9854
was used for calculations of heats of formation on the b3lyp/6 - 31g(@xmath41 ) level of theory as recommended in ref . .
the performance of our fitted functionals will be compared to related and popular density functionals : pbe @xcite , tpss @xcite , revtpss @xcite , m06-l @xcite , mgga_ms0 @xcite , pbeh @xcite , tpssh @xcite , revtpssh @xcite , b3lyp @xcite , blyp @xcite , and m06 @xcite .
calculations of open - shell species were carried out via spin - unrestricted formalisms .
errors are reported as @xmath42 .
the semilocal exchange energy of a spin - unpolarized density can be written as : e_x^sl[n]=d^3 r n _ x^unif(n)f_x(p , ) .
[ eq : ex_sl ] here , @xmath43 is the exchange energy per particle of a ueg with @xmath44 , @xmath45 , and @xmath46 the enhancement factor with @xmath47 for lsda .
the expression for the spin - polarized case follows from the exact spin scaling @xcite . in ref . ,
a simple exchange enhancement factor that disentangles the @xmath22- and @xmath48-dependence was proposed , f_x^int(p , ) = f_x^1(p)+f()[f_x^0(p)-f_x^1(p ) ] , [ eq : fx_mgga ] where @xmath49 and @xmath50 .
@xmath51 interpolates between @xmath52 and @xmath53 through @xmath54 . here , in order to tune the @xmath22 dependence , we add a parameter @xmath55 in the denominator of the original @xmath54 , which reads now @xmath56 . for a slowly varying density where @xmath28 , @xmath54 is controlled by its numerator and
we recover the second order gradient expansion with the first principle coefficient @xmath57 @xcite as in ref . .
the parameter @xmath58 is determined for each @xmath59 to reproduce the exchange energy of the hydrogen atom , where @xmath27 .
since @xmath60 is negligible for @xmath61 and the numerator of @xmath54 modulates the behavior of @xmath51 for @xmath28 , @xmath55 mainly controls the behavior for @xmath62 . on the other hand
, @xmath59 mainly controls the behavior of @xmath51 for large @xmath26 . in the original form
@xcite , @xmath63 , and @xmath64 was determined by the hartree - fock exchange energy of the 12-noninteracting - electron hydrogenic anion with nuclear charge z=1
@xcite . fig . 2 of ref
. showed that the shell regions of the hydrogenic anion are associated with @xmath65 and the intershell regions with @xmath62 , which thus can provide information to guide the exchange functional from @xmath28 to @xmath66 . here , we release the constraint of the hydrogenic anion and rely on some molecular training sets to provide information to determine both @xmath55 and @xmath59 by fitting them to heats of formation and barrier heights .
for the correlation part , we use the variant of the pbe correlation @xcite . when we plug the above exchange and correlation parts into eq .
( [ global_hybrid ] ) , we obtain a global hybrid mgga with three parameters , @xmath59 , @xmath55 , and @xmath33 , which control the dependence of the global hybrid on the density gradient , the kinetic energy density , and the exact exchange , respectively . in the next section ,
we describe the parametrization procedure and the performance of our best functionals .
to get a broad overview about the errors which the semilocal functional class mgga_ms makes using different values of @xmath59 and @xmath55 , we define the average mean absolute error ( amae ) using the mean absolute error ( mae ) values from the ae6 and bh6 test sets : @xmath67 , \end{aligned}\ ] ] and present the plotted surface of amae values in figure [ figure : amae ] .
the figure shows that there are two useful possibilities of fitting a semilocal functional of our mgga_ms form : first , we can design a one - parameter functional by setting the parameter @xmath55 to one and fitting @xmath59 to minimize a chosen error measure .
second , both parameters can be fitted independently .
the second choice will yield a lower amae value .
for the independent fit , we also include the exact exchange admixture parameter @xmath33 into the fitting procedure .
the corresponding semilocal functional will be obtained by setting @xmath33 to zero while keeping the other global hybrid parameters fixed .
variants with errors in kcal / mol shows a broad range of rather small errors ] our chosen error measure contains the mean absolute errors ( mae s ) of the heats of formation of the g2/97 data set and barrier heights of the bh42/03 test sets : @xmath68 .\end{aligned}\ ] ] the resulting parameter sets , me , and mae values are shown in table [ table : parameter_g2bh42 ] and compared to other popular semilocal and hybrid functionals .
mgga_ms0 is the original mgga_ms functional from ref . with no parameter fitting to training sets while mgga_ms1 is our semilocal one - parameter functional .
mgga_ms2h is the new hybrid functional and mgga_ms2 its semilocal complement .
our hybrid mgga_ms2h has a rather low exact exchange admixture of 9% as expected when a good semilocal functional is hybridized @xcite .
comparing the parameter set of mgga_ms2 with the amae surface in figure [ figure : amae ] shows that its parameter set is quite close to the global minimum of the amae , although it has not been refitted after exact exchange admixture was reduced from 9% to 0% .
the functional blyp * is not commonly used , but it is the underlying semilocal functional of b3lyp .
it is included in our tables for comparison purposes only .
as shown in tables [ table : parameter_g2bh42 ] and [ table : g3bh76 ] , blyp * turns the underestimations of blyp for atomization energies into overestimations . from all semilocal functionals in table [ table : parameter_g2bh42 ]
, mgga_ms1 performs best for g2/97 while m06-l performs best for bh42/03 .
mgga_ms0 , mgga_ms1 , and mgga_ms2 functionals are the next best performers for the bh42/03 test set , which behave similarly despite very different parameter sets . for the hybrids
, b3lyp performs best for g2/97 and m06 for bh42/03 .
our global hybrid mgga_ms2h performs similarly to b3lyp and pbeh for the bh42/03 test set and similar to pbeh for the g2/97 test set .
note that pbe , tpss , revtpss , and mgga_ms0 are nonempirical functionals , while the number of empirical parameters as indicated in the parentheses of mgga_ms1 ( 1 ) , mgga_ms2 ( 2 ) , tpssh ( 1 ) , revtpssh ( 1 ) , and mgga_ms2h ( 3 ) are an order of magnitude smaller than those of m06-l and m06 .
mgga_ms1 performs better than mgga_ms2 and mgga_ms0 for all molecular test sets discussed so far .
this and figure [ figure : amae ] indicate that one can construct another mgga_ms functional which performs even better than mgga_ms1 .
however , figure [ figure : amae ] suggests that the improvement over mgga_ms1 will not be significant .
we found that @xmath69 , @xmath70 , and @xmath71 is close to optimal for the error measure @xmath72 , but the minimum in parameter space is rather shallow and broad . with mae values for g2/97 of 4.5
kcal / mol and for bh42/03 of 5.6 kcal / mol , this variant would improve over mgga_ms1 only slightly for g2/97 and not at all for bh42/03 on the expense of an additional empirical parameter .
thus , we drop this variant from now on . [ cols="<,^,^,^,^,^,^,^,^,^,^ " , ] + [ table : s22_d3 ] ref . presents a collection of different techniques for functionals sorted by the way weak interactions are accounted for : step 1 ( dft - d @xcite , etc ) , step 2 ( dft - d3 @xcite , vdw(ts ) @xcite , etc ) , step 3 ( long - range density functionals @xcite ) , and step 4 and higher ( rpa @xcite , etc ) .
representative functionals of each group are tested for the s22 test set in ref . .
note that the basis sets differ compared to our calculations , but the results can be compared roughly .
mgga_ms0 and mgga_ms2 yield results that are close to the vdw - df @xcite which is on step 3 of this classification , while mgga_ms1 fits best to the ground ( below step 1 ) .
it should be noted that the classification is methodological and not based on performance as some functionals in step 1 and 2 outperform some funtionals on step 3 in ref . .
table [ table : s22_d3 ] gives results of the mgga_ms functionals with the d3 correction of grimme _ et al _
the parameters in the d3 correction term were optimized by fitting to the s22 data set for each functional .
we found that the optimization is insensitive with respect to the parameter @xmath73 , and thus set @xmath74 .
then , there is only one fitting parameter @xmath75 left in the d3 correction term .
@xmath75 , which is the scaling factor of cutoff radii and determines the range of the dispersion correction inversely @xcite , is 1.15 , 1.05 , 1.14 , and 1.14 for mgga_ms0 , mgga_ms1 , mgga_ms2 , and mgga_ms2h , respectively .
all the four mgga_ms functionals with the d3 corrections deliver for the s22 data set similar maes around 0.3 kcal / mol , which are among the best of the d3-corrected functionals tested in ref . .
the smaller @xmath75 for mgga_ms1 indicates that mgga_ms1 captures less intermediate - range weak interactions than the other three do , as also shown in table [ table : s22 ] .
table [ table : s22_d3 ] also shows that the d3 corrections have noticeable effects on the formation energies and the barrier heights .
mgga_ms1+d3 is the best among the three mggas , while the hybrid mgga_ms2h+d3 improves further the formation energies and the barrier heights .
the functional pairs of global hybrid and its corresponding semilocal complement perform very similarly for weak interactions on average .
the pair of blyp * and b3lyp is the outlier , which might be related to the fact that b88 is too repulsive for rare gas dimers even when comparing to the exact exchange @xcite .
this general trend suggests that the performance of a hybrid of a semilocal functional inherits its performance for weak interactions from its corresponding pure semilocal functional .
a possible explanation is that weak interactions are mainly correlation effects where exact exchange does not contribute . on the other hand ,
heats of formation , atomization energies , and barrier heights are significantly affected by correlation and exact exchange contributions . however , in our constructions , mgga_ms2 is a byproduct of mgga_ms2h , which is fitted to the atomization energies of the g2/97 test set and the barrier heights of the bh42/03 .
surprisingly , mgga_ms2h performs reasonably well for the s22 set , much better than the other four hybrids , namely , pbeh , tpssh , revtpssh , and b3lyp .
note the improvement of mgga_ms2h over the other four hybrids on the weak interactions is not a result of sacrifice of accuracy for atomization energies and barrier heights as shown in table [ table : g3bh76 ] .
the good performance of mgga_ms2h on the s22 set is then transfered to mgga_ms2 . when considering both tables [ table : g3bh76 ] and [ table : s22 ] for the properties of atomization energies , barrier heights , and weak interactions , our three - parameter global hybrid mgga_ms2h overall is the second best hybrid after m06 that is heavily parameterized .
compared to mgga_ms2h , our semilocal functional mgga_ms1 performs surprisingly well for atomization energies of molecules and barrier heights of chemical reactions although it contains just one empirical parameter .
however , the gain is obtained at the price of descriptions for weak interactions as demonstrated by the comparison with mgga_ms0 in table [ table : s22 ] .
this shows the limit of tuning only the @xmath29dependence of the enhancement factor .
fortunately , by tuning the @xmath30dependence additionally , mgga_ms2 improves over mgga_ms0 for the atomization energies significantly while remaining comparable for the barrier heights and weak interactions .
by taking advantage of the understanding on the kinetic - energy - density dependence of mggas @xcite , we construct a global hybrid meta - generalized gradient approximation ( hybrid meta - gga ) with three empirical parameters ( mgga_ms2h ) , which has robust performances on the molecular systems for the properties of heats of formation , barrier heights , and noncovalent interactions .
the derived underlying mgga_ms2 improves over the original mgga_ms0 @xcite for heats of formation significantly while retaining good performance for barrier heights and weak interactions . by relaxing the parameter @xmath59 of mgga_ms0
, we also obtained the one - parameter functional mgga_ms1 that performs surprisingly well for heats of formations and barrier heights .
most calculations were done at rice university .
this work was supported by the national science foundation under che-1110884 and the welch foundation ( c-0036 ) .
rh thanks the deutsche forschungsgemeinschaft ( dfg grant ha 5711/2 - 1 ) .
js , bx , and jpp acknowledge nsf support under grant no .
dmr-0854769 , and cooperative agreement no .
eps-1003897 , with additional support from the louisiana board of regents .
99 w. kohn and l.j .
sham , phys .
rev . * 140 * , a1133 ( 1965 ) .
parr and w. yang , _ density functional theory of atoms and molecules _
( oxford university press , oxford , 1989 ) .
perdew and s. kurth , in _ a primer in density functional theory _
( springer lecture notes in physics , vol . *
620 * , 2003 ) .
perdew and y. wang , phys . rev .
b * 45 * , 13244 ( 1992 ) .
j. sun , j.p .
perdew and m. seidl , phys . rev .
b * 81 * , 085123 ( 2010 ) .
perdew , k. burke , and m. ernzerhof , phys .
lett . * 77 * , 3865 ( 1996 ) .
becke , j. chem . phys .
* 98 * , 5648 ( 1993 ) .
becke , phys .
a * 38 * , 3098 ( 1988 ) .
stephens , f.j .
devlin , c.f .
chabalowski , and m.j .
frisch , j. phys .
chem . * 98 * , 11623 ( 1994 ) . c. lee , w. yang , and r.g .
parr , phys .
b * 37 * , 785 ( 1988 ) .
k. burke , j. chem .
phys . * 136 * , 150901 ( 2012 ) .
j. p. perdew and k. schmidt , in _ density functional theory and its applications to materials _
, edited by v. e. van doren , c. van alsenoy , and p. geerlings ( american institute of physics , 2001 ) . j. sun , m. marsman , a. ruzsinszky , g. kresse , and j.p .
perdew , phys .
b * 83 * , 121410(r ) ( 2011 ) .
perdew , a. ruzsinszky , g.i .
csonka , o.a .
vydrov , g.e .
scuseria , l.a .
constantin , x. zhou , and k. burke , phys .
lett . * 100 * , 136406 ( 2008 ) .
j. tao , j.p .
perdew , v.n .
staroverov , and g.e .
scuseria , phys .
lett . * 91 * , 146401 ( 2003 ) .
perdew , a. ruzsinszky , g.i .
csonka , l.a .
constantin , and j. sun , phys .
lett . * 103 * , 026403 ( 2009 ) ; _ ibid . _ * 106 * , 179902 ( 2011 ) ( e ) . j. sun , b. xiao , and a. ruzsinszky , j. chem . phys . * 137 * , 051101 ( 2012 ) .
staroverov , g.e .
scuseria , j. tao , and j.p .
perdew , j. chem . phys . * 119 * , 12129 ( 2003 ) . f. furche and j.p .
perdew , j. chem . phys . * 124 * , 044103 ( 2006 ) .
s. kurth , j.p .
perdew , and p. blaha , int .
j. quantum chem . * 75 * , 889 ( 1999 ) .
y. zhao and d.g .
truhlar , j. chem . phys . * 125 * , 194101 ( 2006 ) .
p. hao , j. sun , b. xiao , a. ruzsinszky , g.i .
csonka , j. tao , and j.p .
perdew , j. chem .
. comput . * 10 * , 1021 ( 2012 ) .
madsen , l. ferrighi , and b. hammer , j. phys .
* 1 * , 515 ( 2010 ) . l. ferrighi , g.k.h .
madsen , and b. hammer , j. chem .
phys . * 135 * , 084704 ( 2011 ) .
r. peverati and d.g .
truhlar , j. phys .
lett . * 3 * , 117 ( 2012 ) .
t. van voorhis and g. e. scuseria , j. chem .
* 109 * , 400 ( 1998 ) .
l. goerigk and s. grimme , _ phys .
* 13 * , 6670 ( 2011 ) .
snyder , m. rupp , k. hansen , k - r .
mueller , and k. burke , phys .
lett . * 108 * , 253002 ( 2012 ) .
perdew , v.n .
staroverov , j. tao , and g.e .
scuseria , phys .
a * 78 * , 052513 ( 2008 ) .
m. dion , h. rydberg , e. schrder , d.c .
langreth , and b.i .
lundqvist , phys .
lett . * 92 * , 246401 ( 2004 ) .
a. ruzsinszky , j.p .
perdew , and g.i .
csonka , j. chem .
* 6 * , 127 ( 2010 ) .
q. wu and w. yang , j. chem .
phys . * 116 * , 515 ( 2002 ) .
becke , j. chem .
* 122 * , 064101 ( 2005 ) .
becke , j. chem . phys . *
122 * , 154104 ( 2005 ) .
becke , j. chem . phys . * 123 * , 024101 ( 2005 ) .
becke , j. chem . phys . * 123 * , 154101 ( 2005 ) .
s. grimme , j. comput . chem .
* 27 * , 1787 ( 2006 ) .
s. grimme , j. antony , s. ehrlich , and h. krieg , j. chem . phys . * 132 * , 154104 ( 2010 ) .
a. tkatchenko and m. scheffler , phys .
lett . * 102 * , 073005 ( 2009 ) .
o. a. vydrov and t. van voorhis , phys .
. lett . * 103 * , 063004 ( 2009 ) .
j. klime , d. r. bowler , and a. michaelides , j. phys . : condens .
matter * 22 * , 022201 ( 2010 ) .
dobson , a. white , and a. rubio , phys .
lett . * 96 * , 073201 ( 2006 ) .
a. ruzsinszky , j.p .
perdew , j. tao , g.i .
csonka , and j.m .
pitarke , phys .
( to appear ) .
j. klime and a. michaelides , j. chem .
137 * , 120901 ( 2012 ) .
becke , j. chem . phys . * 98 * , 1372 ( 1993 ) .
perdew , k. burke , and m. ernzerhof , j. chem . phys .
* 105 * , 9982 ( 1996 ) .
csonka , j.p .
perdew , and a. ruzsinszky , j. chem .
* 6 * , 3688 ( 2010 ) .
becke , j. chem . phys . * 107 * , 8554 ( 1997 ) .
y. zhao and d.g .
truhlar , acc .
* 41 * , 157 ( 2008 ) .
vosko , l. wilk , and m. nusair , can .
* 58 * , 1200 ( 1980 ) .
frisch , trucks , schlegel , _ et al .
_ , gaussian development version revision g.01 , gaussian inc . , wallingford , ct ( 2009 ) .
a.d . mclean and g.s .
chandler , j. chem .
phys . * 72 * , 5639 ( 1980 ) .
r. krishnan , j.s .
binkley , r. seeger , and j.a .
pople , j. chem . phys . * 72 * , 650 ( 1980 ) .
lynch , d.g .
truhlar , j. phys .
a * 107 * , 8996 ( 2003 ) ; erratum : * 108 * , 1460 ( 2004 ) .
r. haunschild and w. klopper , theor .
acc . * 131 * , 1112 ( 2012 ) .
curtiss , k. raghavachari , p.c .
redfern , j.a .
pople , j. chem . phys . * 106 * , 1063 ( 1997 ) .
curtiss , p.c .
redfern , k. raghavachari , j.a .
pople , j. chem . phys . * 109 * , 42 ( 1998 ) .
y. zhao , b.j .
lynch , and d. g. truhlar , j. phys .
a * 108 * , 2715 ( 2004 ) .
r. haunschild and w. klopper , j. chem . phys . *
136 * , 164102 ( 2012 ) .
curtiss , k. raghavachari , p.c .
redfern , j.a .
pople , j. chem . phys . * 112 * , 7374 ( 2000 ) . cof@xmath76 was included in the g2/97 and g3/99 test sets following ref . .
p. jureka , j. poner , j. ern , and p. hobza , phys .
* 8 * , 1985 ( 2006 ) .
y. zhao , n. gnzales - garca , and d.g .
truhlar , j. phys .
a * 109 * , 2012 ( 2005 ) ; erratum : * 109 * , 4942 ( 2006 ) .
r. haunschild , b.g .
janesko , and g.e .
scuseria , j. chem . phys . * 131 * , 154112 ( 2009 ) .
s.f . boys and f. bernadi , mol .
* 19 * , 553 ( 1970 ) .
curtiss , p.c .
redfern , k. raghavachari , and j.a .
pople , j. chem .
phys . * 114 * , 108 ( 2001 ) .
staroverov , g.e .
scuseria , j. tao , and j.p .
perdew , j. chem . phys . *
119 * , 12129 ( 2003 ) ; erratum : * 121 * , 11508 ( 2004 ) . y. zhao and d. g. truhlar , theor .
acc . * 120 * , 215 ( 2008 ) .
perdew , s. kurth , a. zupan , and p. blaha , phys .
lett . * 82 * , 2544 ( 1999 ) ; * 82 * , 5179(e ) ( 1999 ) . p.r .
antoniewicz and l. kleinman , phys .
b * 31 * , 6779 ( 1985 ) .
staroverov , g.e .
scuseria , j.p .
perdew , j.tao , and e.r .
davidson , phys .
a * 70 * , 012502 ( 2004 ) a. ruzsinszky , j. sun , b. xiao , and g.i .
csonka , j. chem .
theo . comp .
* 8 * , 2078 ( 2012 ) .
johnson , a.d .
becke , c.d .
sherrill , and g.a .
dilabio , j. chem . phys . * 131 * , 034111 ( 2009 ) .
f. furche , phys .
b * 64 * , 195120 ( 2001 ) .
f.o . kannemann and a.d .
becke , j. chem .
theor . comput . * 5 * , 721 ( 2009 ) . | we present a global hybrid meta - generalized gradient approximation ( meta - gga ) with three empirical parameters , as well as its underlying semilocal meta - gga and a meta - gga with only one empirical parameter .
all of them are based on the new meta - gga resulting from the understanding of kinetic - energy - density dependence [ j. chem .
phys . * 137 * , 051101 ( 2012 ) ] .
the obtained functionals show robust performances on the considered molecular systems for the properties of heats of formation , barrier heights , and noncovalent interactions .
the pair - wise additive dispersion corrections to the functionals are also presented . |
the lhc is expected to directly probe possible new physics beyond the standard model ( sm ) up to a scale of a few tev . while its data should provide answers to several of the major open questions in the present picture of elementary particle physics
, it is important to start examining how this sensitivity can be further extended at a next generation of colliders .
today we have a number of indications that new physics could be of supersymmetric nature .
if this is the case , the lhc will have a variety of signals to discover these new particles and the linear collider ( lc ) will be required to complement the probe of the susy spectrum with detailed measurements .
however , beyond supersymmetry there is a wide range of other scenarios invoking new phenomena at , and beyond , the tev scale .
they are aimed at explaininig the origin of electro - weak symmetry breaking , if there is no light elementary higgs boson , at stabilising the sm , if susy is not realised in nature , or at embedding the sm in a theory of grand unification .
many of such scenarios predict the existence of new particles that would be manifested as rather spectacular resonances in @xmath0 collisions , if the achievable centre - of - mass energy is sufficient .
a high energy lc represents an ideal laboratory for studying this new physics @xcite .
it also retains an indirect sensitivity , through a precision study of the virtual corrections to electro - weak observables , when their mass exceeds the available centre - of - mass energy .
this paper summarises the results of a series of studies aimed at quantifying the potential of a high energy , high luminosity @xmath0 lc in extending to high scales the probe for new physics .
while a significant activity has already addressed the issues related to a tev - class collider , we now review the potential of a multi - tev lc , such as clic .
the analysis of the lep and slc data has provided a significant experience in the extraction of electro - weak observables , optimising their statistical sensitivity and controlling their systematic uncertainties . at larger centre - of - mass energies ,
the relevant @xmath1 cross sections are significantly reduced and the experimental conditions at the interaction region need to be taken into account in validating the anticipated accuracies on the cross section @xmath2 , forward - backward asymmetries @xmath3 and left - right asymmetries @xmath4 determination at @xmath5 = 1 tev - 5 tev . since the two - fermion cross section is of the order of only 10 fb , it is imperative to achieve high luminosity by reducing the beam - spot sizes . in this regime
the beam - beam effects are important and the primary @xmath0 collision is accompained by several @xmath6 interactions .
being mostly confined in the forward regions , this @xmath7 background reduces the polar angle acceptance for quark flavour tagging and dilutes the jet charge separation using jet charge techniques .
these experimental conditions require efficient and robust algorithms to ensure sensitivity to flavour - specific @xmath8 production .
the statistical accuracy for the determination of @xmath2 , @xmath3 and @xmath4 has been studied , for @xmath9 and @xmath10 , taking the clic parameters at @xmath5 = 3 tev .
the simdet parametrised detector simulation has been used and the @xmath6 background , corresponding to 10 overlayed bunch crossings , has been added to @xmath11 , @xmath10 events .
@xmath10 final states have been identified using an algorithm based on the sampling of the decay charged multiplicity of the highly boosted @xmath12 hadrons at clic energies @xcite .
similarly to lep analyses , the forward - backward asymmetry has been extracted from a fit to the flow of the jet charge @xmath13 defined as @xmath14 , where @xmath15 is the particle charge , @xmath16 its momentum , @xmath17 the jet thrust axis and the sum is extended to all the particles in a given jet . here the presence of additional particles , from the @xmath18 background , causes a broadening of the @xmath13 distribution and thus a dilution of the quark charge separation .
the track selection and the value of the power parameter @xmath19 needed to be optimised as a function of the number of overlayed bunch crossings .
the results are summarised in terms of the relative statistical accuracies @xmath20 in table [ tab : res ] .
another important issue is the accuracy on the luminosity determination , that needs to be controlled to 0.5% , or better .
.relative statistical accuracies on electro - weak observables , obtained for 1 ab@xmath21 of clic data at @xmath5 = 3 tev , including the effect of @xmath6 background .
[ cols="<,^ " , ] fermion compositeness or exchange of very heavy new particles can be described in all generality by four - fermion contact interactions @xcite .
these parametrise the interactions beyond the sm by means of an effective scale , @xmath22 , @xmath23 the strength of the interaction is set by convention as @xmath24 and models can be considered by choosing either @xmath25 or @xmath26 as detailed in table ii .
the contact scale @xmath27 can be interpreted as effect of new particles at a mass @xmath28 , @xmath29 . in order to estimate the sensitivity of electro - weak observables to the contact interaction scale @xmath27
, the statistical accuracies discussed in section ii have been assumed for the @xmath30 and @xmath31 final states .
the systematics of the assumed 0.5% include the contributions from model prediction uncertainties .
results are given in terms of the lower limits on @xmath27 which can be excluded at 95% c.l .
, in figure [ fig : ci ] .
it has been verified that , for the channels considered in the present analysis , the bounds for the different @xmath22 are consistent .
high luminosity @xmath0 collisions at 3 tev can probe @xmath27 at scales of 200 tev , and beyond . for comparison , the corresponding results expected for a lc operating at 1 tev are also shown .
beam polarisation represents an important tool in these studies .
first , it improves the sensitivity to new interactions , through the introduction of the left - right asymmetries @xmath32 and the polarised forward - backward asymmetries @xmath33 in the electro - weak fits .
if both beams can be polarised to @xmath34 and @xmath35 respectively , the relevant parameter is the effective polarisation defined as @xmath36 . in addition to the improved sensitivity , the uncertainty on the effective polarization , can be made smaller than the error on the individual beam polarization measurements .
secondly , in the case of a significant deviation from the sm prediction would be observed , @xmath37 and @xmath38 polarization is greatly beneficial to determine the nature of the new interactions .
this has been studied in details for a lc at 0.51.0 tev @xcite and those results also apply , qualitatively , to a multi - tev collider .
extending the sensitivity to new physics beyond the anticipated reach of the lhc , is a prime aim of future colliders . by accurately measuring electro - weak observables , a lc able to achieve @xmath0 collisions at and beyond 1 tev , with high luminosity ,
can indirectly probe scales extending from tens to several hundreds tev .
giudice , r. rattazzi and j.d .
wells , nucl .
b544 * ( 1999 ) 3 ; e. a. mirabelli , m. perelstein and m. e. peskin , phys .
* 82 * ( 1999 ) 2236 ; t. han , j.d .
lykken , r .- j .
zhang , phys .
* d59 * ( 1999 ) 105006 .
see , for example , t.g .
rizzo and j.d .
wells , phys . rev . *
d61 * 016007 ( 2000 ) ; p.
nath and m. yamaguchi , phys . rev . * d60 * ( 1999 ) 116006 ; m. masip and a. pomarol , phys . rev .
* d60 * ( 1999 ) 096005 ; r. casalbuoni , s. de curtis , d. dominici and r. gatto , phys .
b462 * ( 1999 ) 48 ; a. strumia , phys . lett .
* b466 * ( 1999 ) 107 . | extending the sensitivity to new physics beyond the anticipated reach of the lhc is a prime aim of future colliders .
this paper summarises the potential of an @xmath0 linear collider , at and beyond 1 tev , using a realistic simulation of the detector response and the accelerator induced background .
the possible lc energy - luminosity trade - offs offered in probing multi - tev scales for new phenomena with electro - weak observables are also discussed . |
financial support by the austrian science fund ( projects i543-n20 and j3732-n27 ) and the german research foundation within the dach framework is acknowledged .
40ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.94.026803 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop ( ) \doibase http://dx.doi.org/10.1038/43403 [ * * , ( ) ] , @noop * * , ( ) \doibase http://dx.doi.org/10.1063/1.4819397 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1017/s1431927614000610 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) _ _ , @noop ph.d . thesis , ( ) @noop _ _ ( , ) \doibase http://dx.doi.org/10.1016/s0968-4328(99)00112-2 [ * * , ( ) ] link:\doibase 10.1080/13642819908214851 [ * * , ( ) ] , @noop _ _ , advences in electronics and electron physics ( , ) @noop * * , ( ) @noop @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1107/s0365110x57002194 [ * * , ( ) ] @noop _ _ ( , ) \doibase http://dx.doi.org/10.1016/j.ultramic.2013.11.010 [ * * , ( ) ] | transmission electron microscopy has been a promising candidate for mapping atomic orbitals for a long time . here , we explore its capabilities by a first principles approach .
for the example of defected graphene , exhibiting either an isolated vacancy or a substitutional nitrogen atom , we show that three different kinds of images are to be expected , depending on the orbital character . to judge the feasibility of visualizing orbitals in a real microscope , the effect of the optics aberrations is simulated .
we demonstrate that , by making use of energy - filtering , it should indeed be possible to map atomic orbitals in a state - of - the - art transmission electron microscope .
the possibility to _ see _ atomic orbitals has always attracted great scientific interest . at the same time , however , the real meaning of `` measuring orbitals '' has been a subject that scientists have long and much dwelt upon ( see , e.g. , @xcite and references therein ) . in the past , significant efforts have been devoted to the development of experimental approaches and theoretical models that allow for orbital reconstruction from experimental data @xcite .
based on the generation of higher harmonics by femtoseconds laser pulses , a tomographic reconstruction of the highest occupied molecular orbital ( homo ) for simple diatomic molecules in the gas phase was proposed @xcite .
direct imaging of the homo and the lowest - occupied molecular orbital ( lumo ) of pentacene on a metallic substrate was theoretically predicted and experimentally verified with scanning - tunnelling microscopy ( stm ) @xcite .
more recently , real - space reconstruction of molecular orbitals from angle - resolved photoemission data has been demonstrated @xcite .
this method has been subsequently further developed to retrieve both the spatial distribution @xcite and the phase of electron wavefunctions of pentacene and perylene-3,4,9,10-tetracarboxylic dianhydride ( ptcda ) adsorbed on silver @xcite .
the reconstruction of charge densities and chemical bonds using transmission electron microscopy ( tem ) has been considered @xcite , but only recently the possibility of probing selected transitions to specific unoccupied orbitals by using energy - filtered tem ( eftem ) was demonstrated theoretically .
a first example for the capability of this approach was provided with the oxygen k - edge of rutile tio@xmath0 @xcite .
however the interpretation of experimental tem images for systems like rutile would be complicated because of the multiple elastic scattering of electrons that occurs in thick samples . in this letter , we suggest defective graphene @xcite as the prototypical two dimensional ( 2d ) material to demonstrate the possibility of mapping atomic orbitals using eftem .
we break the ideal @xmath1 hybridization by introducing two different kinds of defects , namely a single isolated vacancy and a substitutional nitrogen atom .
this lifts the degeneracy of the @xmath2-states , inducing strong modifications to the electronic properties compared to the pristine lattice @xcite . by selecting certain scattering angles ,
dipole - allowed transitions dominate the electron energy loss spectroscopy ( eels ) signal @xcite .
a single - particle description can be safely adopted , since many - body effects do not play a major role in the excitation process .
overall , tem images of these systems can be interpreted in terms of bare @xmath3 transitions . in an eftem experiment ,
an incoming beam of high - energy electrons ( of the order of 100 kev ) is directed to the target where it scatters at the atoms either elastically or inelastically .
the outgoing electron beam is detected and analyzed .
state of the art image simulations generally only include elastic scattering of the electrons using the multi - slice approach @xcite . in the case of eftem for a thin sample ,
the influence of elastic scattering becomes negligible , and inelastic scattering gives the dominant contribution to the formation of the images . the key quantity to describe the inelastic scattering of electrons , which is probed by eels , is the mixed dynamic form factor ( mdff ) .
it can be interpreted as a weighted sum of transition matrix elements between initial and final states @xmath4 and @xmath5 of the target electron @xcite : @xmath6 with energies @xmath7 and @xmath8 .
@xmath9 is the energy - loss of the fast electron of the incident beam , @xmath10 and @xmath11 are the wavevectors of the perturbing and induced density fluctuations , respectively .
if many - body effects can be neglected , this picture can be simplified for dipole - allowed transitions . in this case , using the spherical harmonics as basis for the target states and referring to transitions originating from a single state ( as in @xmath3 excitations ) , the mdff is @xcite @xmath12 where @xmath13 are spherical harmonics , @xmath14 is an integral of the spherical bessel function @xmath15 weighted over the initial and final radial wavefunctions . @xmath16 and @xmath17 indicate the azimuthal and magnetic quantum number of the final state of the target electron , and @xmath18 and @xmath19 are the angular momenta transferred during the transition .
@xmath20 is a quantity that describes crystal - field effects and is proportional to the cross - density of states ( xdos ) @xmath21 where @xmath22 is the angular part of the final wave function , @xmath23 is the band index , and @xmath24 is a k - point in the first brillouin zone . compared to the density of states ( dos ) , the xdos includes also non - diagonal terms connecting states with different angular momenta . as @xmath25 is a hermitian matrix @xcite , the mdff can be diagonalized .
therefore , assuming that the target s final states of the @xmath3 excitation are not degenerate , the transition matrix elements reflect the azimuthal shape of the final single - particle states and can thus be separated by using energy - filtering .
ground - state calculations are performed using density - functional theory and the full - potential augmented planewave plus local - orbital method , as implemented in exciting @xcite . introducing a vacancy or a substitutional atom , a 5@xmath265 supercell
is set up , hosting 49 and 50 atoms , respectively ( fig . 1 and 2 in the supplemental material @xcite ) . the space group and thus the number of inequivalent carbon atoms ( 13 ) is the same in both cases .
we adopt a lattice parameter of @xmath27=4.648 bohr , corresponding to a bond length of 2.683 bohr , while the cell size perpendicular to the graphene plane is set to c=37.794 bohr in order to prevent interactions between the periodically repeated layers .
exchange - correlation effects are treated by the pbe functional @xcite .
the brillouin zone is sampled with an 8@xmath268@xmath261 k - point grid .
the structures are relaxed down to a residual force lower than 0.0005 ha / bohr acting on each atom .
interatomic distances between atoms of the relaxed structures , up to the seventh nearest neighbor , are given in table [ tab1 ] .
upon relaxation , the atoms surrounding the vacancy move slightly away from it , thus shortening the bond lengths with the next nearest neighbors , @xmath28 , compared to the unperturbed system .
the effect of the vacancy extends up to the fourth neighbours , whereas it is almost negligible for more distant atoms ( more information about the relaxed structures can be found in the supplemental material @xcite ) . in the case of nitrogen doping , the substitutional atom does not strongly influence the atomic configuration of the system .
this happens because the nitrogen - carbon bond length is just slightly shortened with respect to the carbon - carbon bond length in pristine graphene . for all the systems , we have investigated dipole - allowed transitions at the k - edge of carbon , assuming an incoming electron beam perpendicular to the graphene plane .
.[tab1 ] bond lengths between atoms up to the seventh nearest neighbor , @xmath29 , @xmath28 , @xmath30 , @xmath31 , @xmath32 , and @xmath33 , for graphene doped with nitrogen ( top row ) and with a vacancy ( bottom row ) .
@xmath34 are the relative deviations from those of pristine graphene .
@xmath35 indicates the defect site .
( [ cols="<,^,^,^,^,^,^,^",options="header " , ] in fig .
[ fig_dos ] , the projected density of states ( pdos ) of pristine graphene ( upper panel ) and of the first nearest - neighbor atom for nitrogen - doped graphene ( middle panel ) and graphene with a single vacancy ( bottom ) , respectively , is plotted for empty states up to 12 ev above the fermi energy . here , @xmath36 , @xmath37 , and @xmath38 represent the local cartesian coordinates at the individual atomic sites as determined by the point - group symmetry .
in particular , @xmath38 is the axis perpendicular to the graphene , i.e. , ( @xmath36 , @xmath37 ) plane .
all the other atoms of the defective systems exhibit a pdos with very similar character as in pristine graphene , besides the second and third nearest neighbors which are slightly affected by the defect @xcite .
local projected density of states ( pdos ) of carbon in pristine graphene ( upper panel ) , and of the first nearest - neighbor atom for nitrogen - doped graphene ( middle panel ) and graphene with a single vacancy ( bottom ) .
@xmath39 , @xmath40 , and @xmath41 states are indicated by the red , green , and black lines , respectively .
note that , in the lower panel , the peak close to the fermi energy exceeds the boundary of the box , with a maximum at about 2 states / ev . in the case of pristine graphene , the red line indicates the sum of @xmath42 and @xmath43 .
the three colored areas labeled as a , b , and c indicate the energy ranges for which the tem images have been calculated.,scaledwidth=48.0% ] in pristine graphene , antibonding @xmath44 and @xmath45 states are clearly recognizable at about 2 ev and 9 ev , respectively .
as already reported in literature @xcite , the introduction of a vacancy or a substitutional nitrogen has a significant influence on the electronic structure .
a consequence of the doping atom is lifting the degeneracy of @xmath39 and @xmath40 that is significant for the first nearest neighbors ( middle panel in fig .
[ fig_dos ] ) .
this effect is particularly evidenced by the appearance of bands at about 5 ev , which exhibit @xmath40 character . here ,
three different regions can be easily identified : a ) from 0 ev to 4 ev , the bands have only @xmath41 character ; energy ranges that present such dos character will be referred to as @xmath46 .
b ) for energies higher than 6 ev , there are contributions from @xmath39 , @xmath40 , and @xmath41 .
the only difference to ideal graphene is the lifted degeneracy of @xmath39 and @xmath40 .
this defines a new kind of region , named @xmath47 .
c ) between 4 ev and 6 ev , the @xmath39 character of the first nearest neighbor is much less pronounced than that of @xmath40 , while all the other atoms have only @xmath41 character ; this region will be referred to as @xmath48 . in the case of graphene with a vacancy
, the same kinds of regions can be identified , but corresponding to different energy ranges . here , the @xmath49 type is found between 0.5 ev and 7 ev ; @xmath47 encompasses energies above 7 ev ; @xmath48 is a small energy window , just few tenths of ev close to the fermi energy .
we find similar kinds of dos characters also for damaged nitrogen - doped graphene , i.e. , graphene with a substitutional nitrogen and a vacancy located near it ; such defects have beed recently reported in tem measurements of nitrogen - doped graphene @xcite .
details of this calculation and the corresponding simulated tem images can be found in the supplemental material @xcite . to investigate the impact of the local electronic structure ( pdos ) on the eftem images , we first consider the ideal case of a perfect microscope with an acceleration voltage of 300 kev . in this case , the recorded images correspond to the intensity of the exit wavefunction in the multislice simulation @xcite .
the finite resolution of the spectrometer is taken into account by simulating images every 0.05 ev in 2 ev - broad energy ranges ( regions a , b , and c in fig .
[ fig_theo ] ) and then summing them up to get the final images .
each image is shown in contrast - optimized grayscale .
simulated real - space intensity of the electron s exit wave function after propagation of an incident planewave through a graphene layer in presence of a nitrogen substitutional atom ( upper panels ) and a vacancy ( bottom panels ) .
the colored lines indicate the energy regions highlighted in fig .
[ fig_dos].,scaledwidth=48.0% ] first , we analyze graphene with nitrogen doping ( fig . [ fig_theo ] , upper panels ) . here , in the region close to the fermi level ( a in fig .
[ fig_dos ] ) there are only contributions from @xmath41 orbitals . the image is then formed by disk - like features where their arrangement clearly visualizes the missing atoms ( upper left panel in fig . [ fig_theo ] ) . at an energy loss between 4 and 6 ev above the carbon k - edge
( region b in fig .
[ fig_dos ] ) , there is a @xmath48-like region . we expect to see contributions from @xmath40 of the atom closest to the nitrogen , but no ( or very little ) signal coming from the other atoms .
this happens because @xmath40 lies on a plane perpendicular to the incoming electron beam and its magnitude is more intense than the one of @xmath41 ; thus its contribution to the final signal overcomes the one from @xmath41 states .
consequently , only the @xmath40 orbitals of the three atoms surrounding the nitrogen are visible , which are pointing towards the defect , as imposed by the local @xmath50 symmetry ( upper middle panel in fig .
[ fig_theo ] ) . at an excitation energy between 8 and 10 ev above the k - edge ( region c in fig .
[ fig_dos ] ) , instead , there are contributions from all the @xmath2 states . since , however , @xmath39 and @xmath40 lie in a plane perpendicular to the beam axis , their contribution to the final signal dominates over the one from @xmath41 states . as a consequence ,
the image is composed of ring - like features , stemming solely from @xmath39 and @xmath40 states , arranged in hexagons ( upper right panel in fig .
[ fig_theo ] ) . due to symmetry
breaking , the intensity is not uniform , neither along a ring ( since @xmath39 and @xmath40 states are non - degenerate ) nor among different rings ( due to non - equivalent atomic sites ) . the corresponding images for the system with a vacancy ( bottom panels in fig .
[ fig_theo ] ) appear nearly identical to the ones above , but at different energy ranges .
this can be understood by comparing the pdos of the two systems . between 0 and 2 ev ,
for instance , we have a @xmath49-like region in the case of nitrogen - doped graphene , and both @xmath48 and @xmath49 in the case of graphene with a vacancy .
due to the similarity of the two systems , we will , in the following , focus on doped graphene and show the corresponding analysis for graphene with a single vacancy in the supplemental material @xcite . in order to predict the outcome of real experiments , we now visualize the effect of the optics aberrations and of a more realistic acceleration voltage on these images .
we have simulated an electron - beam acceleration voltage of 80 kev and the operating parameters of two different kinds of microscopes , the fei tecnai g@xmath51 f20 and fei titan g@xmath51 60 - 300 .
the former has a spherical aberration @xmath52 = 1.2 mm , corresponding to a an extended scherzer defocus of 849 @xmath53 , while the latter is a last - generation aberration - corrected microscope , i.e. , exhibiting no spherical and chromatic aberrations . in view of that
, chromatic aberrations are not included in the calculations .
the images corresponding to the energy regions a , b , and c are shown in fig .
[ fig_exp_n ] . simulated tem image of nitrogen - doped graphene .
the coloured lines indicate the energy regions highlighted in fig .
[ fig_dos ] .
80 kev incident beam energy and lenses as in a titan ( upper panels ) and a tecnai microscope ( bottom panels ) were assumed.,scaledwidth=48.0% ] because of the lower resolution of the tecnai microscope , all the features are blurred ( lower panels ) compared to the ideal images . therefore ,
neither the atomic positions , nor the orbital shapes can be retrieved from them . on the other hand , images simulated by taking into account the aberration - corrected optics of the titan microscope are very sharp and let us identify all the features already observed for the idealized situation previously described .
this can be easily seen , comparing the upper panels of figs .
[ fig_exp_n ] and [ fig_theo ] .
in particular , at an energy loss of 5 ev ( region b ) , the @xmath40 orbitals are visible , as in the ideal images .
this clearly demonstrates the potential ability of aberration - corrected microscopes to visualize atomic orbitals with eftem , especially in a system like graphene .
this conclusion also holds when considering noise caused by the finite electron dose ( see supplemental material @xcite for corresponding images ) . in summary , we have predicted the possibility of performing orbital mapping in low - dimension systems using eftem and we have demonstrated it with the prototypical example of defective graphene . in particular , we have shown that , as far as the optics is concerned , reasonable image resolution may already nowadays be experimentally achievable with last generation aberration - corrected microscopes like a fei titan g@xmath51 60 - 300 and even more with improved instruments of the next generation .
however , additional work is necessary to reduce artifacts such as noise , drift , instabilities and damage .
the inelastic cross section for the carbon k - edge ionisation is about a factor of 10 smaller than the elastic scattering cross section on a carbon atom @xcite .
the intensity collected within an energy window of 2 ev as in fig .
[ fig_theo ] is @xmath545% of the total k - edge intensity .
so , order - of - magnitude - wise , in order to obtain the same snr as in elastic imaging , we need at least 200 times more incident dose which means a dwell time of the order of several minutes for last generation tems in eftem mode .
there is no fundamental law that would forbid such an experiment with today s equipment ; however , it is hampered by drift ( which must be well below the interatomic distance during the exposure time ) , instabilities , and radiation damage .
a new route to circumvent radiation damage based on an eftem low - dose technique was proposed recently @xcite .
this may solve the problem in future .
we have identified three different kinds of images that are expected to be acquired in an eftem experiment , depending on the character of the dos : when only @xmath41 states are present in the electronic structure , the corresponding images are composed of disk - like features .
when the dos is characterized by contributions from all @xmath2 states , ring - like features are seen that , however , only originate from a convolution of @xmath39 and @xmath40 states , while the @xmath55 character is not visible .
when the @xmath40 character strongly exceeds the one of @xmath39 , only a single orbital is recorded .
we expect this work to trigger new experiments on defective graphene and similar systems . |
we consider the problem of encoding and decoding binary codes of constant hamming weight @xmath0 and block length @xmath1 .
such codes are useful in a variety of applications : a few examples are fault - tolerant circuit design and computing @xcite , pattern generation for circuit testing @xcite , identification coding @xcite and optical overlay networks @xcite .
the problem of interest is that of designing the encoder and decoder , i.e. , the problem of mapping all binary ( information ) vectors of a given length onto a subset of length-@xmath1 vectors of constant hamming weight @xmath0 in a one - to - one manner . in this work ,
we propose a novel geometric method in which information and code vectors are represented by vectors in @xmath0-dimensional euclidean space , covering polytopes for the two sets are identified , and a one - to - one mapping is established by dissecting the covering polytopes in a specific manner .
this approach results in an invertible integer - to - integer mapping , thereby ensuring unique decodability .
the proposed algorithm has a natural recursive structure , and an inductive proof is given for unique decodability .
the issue of efficient encoding and decoding is also addressed .
we show that the proposed algorithm has complexity @xmath2 , where @xmath0 is the weight of the codeword , independent of the codeword length .
dissections are of considerable interest in geometry , partly as a source of puzzles , but more importantly because they are intrinsic to the notion of volume . of the @xmath3 problems posed by david hilbert at the international congress of mathematicians in 1900 , the third problem dealt with dissections .
hilbert asked for a proof that there are two tetrahedra of the same volume with the property that it is impossible to dissect one into a finite number of pieces that can be rearranged to give the other , i.e. , that the two tetrahedra are not equidecomposable .
the problem was immediately solved by dehn @xcite . in 1965 , after @xmath4 years of effort , sydler @xcite completed dehn s work .
the dehn - sydler theorem states that a necessary and sufficient condition for two polyhedra to be equidecomposable is that they have the same volume and the same dehn invariant .
this invariant is a certain function of the edge - lengths and dihedral angles of the polyhedron .
an analogous theorem holds in four dimensions ( jessen @xcite ) , but in higher dimensions it is known only that equality of the dehn invariants is a necessary condition . in two dimensions
any two polygons of equal area are equidecomposable , a result due to bolyai and gerwein ( see boltianskii @xcite ) . among other books dealing with the classical dissection problem in two and three dimensions we mention in particular frederickson @xcite , lindgren @xcite and sah @xcite .
the remainder of the paper is organized as follows .
we provide background and review relevant previous work in section [ sec : review ] .
section [ sec : geo ] describes our geometric approach and gives some low - dimensional examples .
encoding and decoding algorithms are then given in section [ sec : alg ] , and the correctness of the algorithms is established .
section [ sec : con ] summarizes the paper .
let us denote the hamming weight of a length-@xmath1 binary sequence @xmath5 by @xmath6 , where @xmath7 is the cardinality of a set .
an @xmath8 constant weight binary code @xmath9 is a set of length-@xmath1 sequences such that any sequence @xmath10 has weight @xmath11 . if @xmath12 is an @xmath8 constant weight code , then its rate @xmath13 . for fixed @xmath14 and @xmath15
, we have @xmath16 where @xmath17 is the entropy function .
thus @xmath18 is maximized when @xmath19 , i.e. , the asymptotic rate is highest when the code is balanced .
the ( asymptotic ) efficiency of a code relative to an infinite - length code with the same weight to length ratio @xmath20 , given by @xmath21 , can be written as @xmath22 where @xmath23 and @xmath24 .
the first term , @xmath25 , is the efficiency of a particular code relative to the best possible code with the same length and weight ; the second term , @xmath26 , is the efficiency of the best finite - length code relative to the best infinite - length code . from stirling s formula
we have @xmath27 a plot of @xmath26 as a function of @xmath1 is given in fig .
[ fig : efficiency ] for @xmath19 .
the slow convergence visible here is the reason one needs codes with large block lengths . as a function of block length
when @xmath19,width=283 ] [ fig1 ] comprehensive tables and construction techniques for binary constant weight codes can be found in @xcite and the references therein . however , the problem of finding efficient encoding and decoding algorithms has received considerably less attention .
we briefly discuss two previous methods that are relevant to our work .
the first , a general purpose technique based on the idea of lexicographic ordering and enumeration of codewords in a codebook ( schalkwijk @xcite , cover @xcite ) is an example of ranking / unranking algorithms that are well studied in the combinatorial literature ( nijenhuis and wilf @xcite ) .
we refer to this as the _ enumerative _ approach .
the second ( knuth @xcite ) is a special - purpose , highly efficient technique that works for balanced codes , i.e. , when @xmath28 , and is referred to as the _ complementation _ method .
the enumerative approach orders the codewords lexicographically ( with respect to the partial order defined by @xmath29 ) , as in a dictionary .
the encoder computes the codeword from its dictionary index , and the decoder computes the dictionary index from the codeword .
the method is effective because there is a simple formula involving binomial coefficients for computing the lexicographic index of a codeword .
the resulting code is fully efficient in the sense that @xmath30 .
however , this method requires the computation of the exact values of binomial coefficients @xmath31 , and requires registers of length @xmath32 , which limits its usefulness .
an alternative is to use arithmetic coding ( rissanen and langdon @xcite , rissanen @xcite ; see also cover and thomas @xcite ) .
arithmetic coding is an efficient variable length source coding technique for finite alphabet sources .
given a source alphabet and a simple probability model for sequences @xmath33 , let @xmath34 and @xmath35 denote the probability distribution and cumulative distribution function , respectively .
an arithmetic encoder represents @xmath33 by a number in the interval @xmath36 $ ] .
the implementation of such a coder can also run into problems with very long registers , but elegant finite - length implementations are known and are widely used ( witten , neal and cleary @xcite ) . for constant weight codes , the idea is to reverse the roles of encoder and decoder , i.e. , to use an arithmetic decoder as an encoder and an arithmetic encoder as a constant weight decoder ( ramabadran @xcite ) .
ramabadran gives an efficient algorithm based on an adaptive probability model , in the sense that the probability that the incoming bit is a @xmath37 depends on the number of @xmath37 s that have already occurred .
this approach successfully overcomes the finite - register - length constraints associated with computing the binomial coefficients and the resulting efficiency is often very high , in many cases the loss of information being at most one bit . the encoding complexity of the method is @xmath32 .
knuth s complementation method @xcite relies on the key observation that if the bits of a length-@xmath1 binary sequence are complemented sequentially , starting from the beginning , there must be a point at which the weight is equal to @xmath38 .
given the transformed sequence , it is possible to recover the original sequence by specifying how many bits were complemented ( or the weight of the original sequence ) .
this information is provided by a ( relatively short ) constant weight check string , and the resulting code consists of the transformed sequence followed by the constant weight check bits . in a series of papers , bose and
colleagues extended knuth s method in various ways , and determined the limits of this approach ( see @xcite and references therein ) .
the method is simple and efficient , and even though the overall complexity is @xmath32 , for @xmath39 we found it to be eight times as fast as the method based on arithmetic codes .
however , the method only works for balanced codes , which restricts its applicability .
the two techniques that we have described above both have complexity that depends on the length @xmath1 of the codewords .
in contrast , the complexity of our algorithm depends only on the weight @xmath0 , which makes it more suitable for codes with relatively low weight . as a final piece of background information
, we define what we mean by a dissection .
we assume the reader is familiar with the terminology of polytopes ( see for example coxeter @xcite , grnbaum @xcite , ziegler @xcite ) . two polytopes @xmath40 and @xmath41 in @xmath42
are said to be _ congruent _ if @xmath41 can be obtained from @xmath40 by a translation , a rotation and possibly a reflection in a hyperplane .
two polytopes @xmath40 and @xmath41 in @xmath42 are said to be _
equidecomposable _ if they can be decomposed into finite sets of polytopes @xmath43 and @xmath44 , respectively , for some positive integer @xmath45 , such that @xmath46 and @xmath47 are congruent for all @xmath48 ( see frederickson @xcite ) .
that is , @xmath40 is the disjoint union of the polytopes @xmath46 , and similarly for @xmath41 .
if this is the case then we say that @xmath40 can be _ dissected _ to give @xmath41 ( and that @xmath41 can be dissected to give @xmath40 ) .
note that we allow reflections in the dissection : there are at least four reasons for doing so .
( i ) it makes no difference to the _ existence _ of the dissection , since if two polytopes are equidecomposable using reflections they are also equidecomposable without using reflections .
this is a classical theorem in two and three dimensions ( * ? ? ?
* chap . 20 ) and the proof is easily generalized to higher dimensions .
( ii ) when studying congruences , it is simpler not to have to worry about whether the determinant of the orthogonal matrix has determinant @xmath49 or @xmath50 .
( iii ) allowing reflections often reduces the number of pieces .
( iv ) since our dissections are mostly in dimensions greater than three , the question of `` physical realizability '' is usually irrelevant .
note also that we do not require that the @xmath46 can be obtained from @xmath40 by a succession of cuts along infinite hyperplanes .
all we require is that @xmath40 be a disjoint union of the @xmath46 .
one final technical point : when defining dissections using coordinates , as in eqns .
( [ eq2da ] ) , ( [ eq2db ] ) below , we use a mixture of @xmath51 and @xmath52 signs in order to have unambiguously defined maps .
this is essential for our application . on the other hand ,
it means that the `` pieces '' in the dissection may be missing certain boundaries .
it should therefore be understood that if we were focusing on the dissections themselves , we would replace each piece by its topological closure . for further information about dissections
see the books mentioned in section [ sec : intro ] .
in this section , we first consider the problem of encoding and decoding a binary constant weight code of weight @xmath53 and arbitrary length @xmath1 , i.e. , where there are only two bits set to @xmath37 in any codeword . our approach is based on the fact that vectors of weight two can be represented as points in two - dimensional euclidean space , and can be scaled , or normalized , to lie in a right triangle .
this approach is then extended , first to weight @xmath54 , and then to arbitrary weights @xmath0 . for any weight @xmath0 and block length @xmath1 ,
let @xmath55 denote the set of all weight @xmath0 vectors , with @xmath56 .
our codebook @xmath9 will be a subset of @xmath55 , and will be equal to @xmath55 for a fully efficient code , i.e. , when @xmath30 .
we will represent a codeword by the @xmath0-tuple @xmath57 , @xmath58 , where @xmath59 is the position of the @xmath60th @xmath37 in the codeword , counting from the left .
if we normalize these indices @xmath59 by dividing them by @xmath1 , the codebook @xmath9 becomes a discrete subset of the polytope @xmath61 , the convex hull of the points @xmath62 .
@xmath63 is a right triangle , @xmath64 is a right tetrahedron and in general we will call @xmath61 a _ unit orthoscheme_-dimensional simplex having an edge path consisting of @xmath0 totally orthogonal vectors ( coxeter @xcite ) . in a _ unit orthoscheme _ these edges all have length @xmath37 . ] .
the set of inputs to the encoder will be denoted by @xmath65 : we assume that this consists of @xmath0-tuples @xmath66 which range over a @xmath0-dimensional hyper - rectangle or `` brick '' .
after normalization by dividing the @xmath67 by @xmath1 , we may assume that the input vector is a point in the hyper - rectangle or `` brick '' @xmath68 we will use @xmath69 and @xmath70 to denote the normalized versions of the input vector and codeword , respectively , defined by @xmath71 and @xmath72 for @xmath73 . the basic idea underlying our approach is to find a dissection of @xmath74 that gives @xmath61 .
the encoding and decoding algorithms are obtained by tracking how the points @xmath75 and @xmath76 move during the dissection .
the volume of @xmath74 is @xmath77 .
this is also the volume of @xmath61 , as the following argument shows .
classify the points @xmath78 in the unit cube @xmath79^w$ ] into @xmath80 regions according to their order when sorted ; the regions are congruent , so all have volume @xmath81 , and the region where the @xmath82 are in nondecreasing order is @xmath61 . to give triangle @xmath63 .
piece 1 may be rotated about center into its new position , or reflected in main diagonal and translated downwards . ,
width=340 ] [ fig2 ] we now return to the case @xmath53 .
there are many ways to dissect the rectangle @xmath83 into the right triangle @xmath63 .
we will consider two such dissections , both two - piece dissections based on fig .
[ fig:2d ] . in the first dissection ,
the triangular piece marked 1 in fig .
[ fig:2d ] is rotated clockwise about the center of the square until it reaches the position shown on the right in fig .
[ fig:2d ] . in the second dissection ,
the piece marked 1 is first reflected in the main diagonal of the square and then translated downwards until it reaches the position shown on the right in fig .
[ fig:2d ] . in both dissections
the piece marked 2 is fixed .
the two dissections can be specified in terms of coordinates and @xmath52 signs , see the remark at the end of section [ sec : review ] . ] as follows .
for the first dissection , we set @xmath84 and for the second , we set @xmath85 the first dissection involves only a rotation , but seems harder to generalize to higher dimensions .
the second one is the one we will generalize ; it uses a reflection , but as mentioned at the end of section [ sec : review ] , this is permitted by the definition of a dissection .
we next illustrate how these dissections can be converted into encoding algorithms for constant weight ( weight @xmath86 ) binary codes .
again there may be several solutions , and the best algorithm may depend on arithmetic properties of @xmath1 ( such as its parity ) .
we work now with the unnormalized sets @xmath87 and @xmath88 . in each case
the output is a weight-@xmath86 binary vector with @xmath37 s in positions @xmath89 and @xmath90 . 1 .
the input is an information vector @xmath91 with @xmath92 and @xmath93 .
2 . if @xmath94 , we set @xmath95 , @xmath96 , otherwise we set @xmath97 and @xmath98 . for @xmath1 even , this algorithm generates all possible @xmath99 codewords . for @xmath1 odd
it generates only @xmath100 codewords , leading to a slight inefficiency , and the following algorithm is to be preferred . 1 .
the input is an information vector @xmath91 with @xmath101 , @xmath102 .
2 . if @xmath94 , we set @xmath95 , @xmath96 , otherwise we set @xmath103 , @xmath104 . for @xmath1 odd , this algorithm generates all @xmath99 codewords , but for @xmath1 even it generates only @xmath99 codewords , again leading to a slight inefficiency . 1 .
the input is an information vector @xmath91 with @xmath92 and @xmath93 .
2 . if @xmath94 , we set @xmath95 , @xmath96 , otherwise we set @xmath105 , @xmath106 . for @xmath1 even , this algorithm generates all @xmath99 codewords , but for @xmath1 odd it generates only @xmath100 codewords , leading to a slight inefficiency .
there is a similar algorithm , not given here , which is better when @xmath1 is odd .
note that only one test is required in any of the encoding algorithms .
the mappings are invertible , with obvious decoding algorithms corresponding to the inverse mappings from @xmath88 to @xmath87 we now extend this method to weight @xmath54 . fortunately , the dehn invariants for both the brick @xmath107 and our unit orthoscheme @xmath64 , which is the tetrahedron with vertices @xmath108 and @xmath109 , are zero ( since in both cases all dihedral angles are rational multiples of @xmath110 ) , and so by the dehn - sydler theorem the polyhedra @xmath107 and @xmath64 _ are _ equidecomposable . as already mentioned in section [ sec : intro ] , the dehn - sydler theorem applies only in three dimensions .
but it will follow from the algorithm given in the next section that @xmath74 and @xmath61 are equidecomposable in all dimensions
. we will continue to describe the encoding step ( the map from @xmath74 to @xmath61 ) first .
we will give an inductive dissection ( see fig .
[ fig:3d ] ) , transforming @xmath107 to @xmath64 in two steps , effectively reducing the dimension by one at each step . in the first step ,
the brick @xmath107 is dissected into a triangular prism ( the product of a right triangle , @xmath63 , and an interval ) , and in the second step this triangular prism is dissected into the tetrahedron @xmath64 .
note that the first step has essentially been solved by the dissection given in eqn .
( [ eq2db ] ) .
[ fig3 ] for the second step we use a four - piece dissection of the triangular prism to the tetrahedron @xmath64 .
this dissection , shown with the tetrahedron and prism superimposed in fig .
[ fig:3dpiece ] , appears to be new .
there is a well - known dissection of the same pair of polyhedra that was first published by hill in 1896 @xcite .
this also uses four pieces , and is discussed in several references : see boltianskii @xcite , cromwell @xcite , frederickson ( * ? ? ? * fig .
20.4 ) , sydler @xcite , wells @xcite . however , hill s dissection seems harder to generalize to higher dimensions .
hill s dissection does have the advantage over ours that it can be accomplished purely by translations and rotations , whereas in our dissection two of the pieces ( pieces labeled 2 and 3 in fig .
[ fig:3dpiece ] ) are also reflected .
however , as mentioned at the end of section [ sec : review ] , this is permitted by the definition of a dissection , and is not a drawback for our application . apart from this , our dissection is simpler than hill s , in the sense that his dissection requires a cut along a skew plane ( @xmath111 ) , whereas all our cuts are parallel to coordinate axes .
[ fig4 ] to obtain the four pieces shown in fig .
[ fig:3dpiece ] , we first make two horizontal cuts along the planes @xmath112 and @xmath113 , dividing the tetrahedron into three slices .
we then cut the middle slice into two by a vertical cut along the plane @xmath114 .
there appears to be a tradition in geometry books that discuss dissections of not giving coordinates for the pieces . to an engineer
this seems unsatisfactory , and so in table [ t1 ] we list the vertices of the four pieces in our dissection .
piece 1 has four vertices , while the other three pieces each have six vertices .
( in the hill dissection the numbers of vertices of the four pieces are @xmath115 , @xmath116 , @xmath117 and @xmath117 respectively . ) given these coordinates , it is not difficult to verify that the four pieces can be reassembled to form the triangular prism , as indicated in fig .
[ fig:3dpiece ] .
as already remarked , pieces 2 and 3 are also reflected ( or `` turned over '' in a fourth dimension ) .
the correctness of the dissection also follows from the alternative description of this dissection given below .
@xmath118 , [ 0,0,1/3 ] , [ 0,1/3,1/3 ] , [ 1/3 , 1/3,1/3 ] .
\\ 2 & [ 0,0,1/3 ] , [ 0,1/3,1/3 ] , [ 1/3,1/3,1/3 ] , \\ & [ 0,0,2/3 ] , [ 0,1/3,2/3 ] , [ 1/3,1/3,2/3 ] .
\\ 3 & [ 0,1/3,1/3 ] , [ 1/3,1/3,1/3 ] , [ 0 , 1/3,2/3 ] , \\ & [ 0,2/3,2/3 ] , [ 2/3,2/3,2/3 ] , [ 1/3,1/3,2/3 ] . \\ 4 & [ 0,0,2/3 ] , [ 0,2/3,2/3 ] , [ 2/3 , 2/3,2/3 ] , \\ &
[ 0,0,1 ] , [ 0,1,1 ] , [ 1 , 1,1 ] .
\\ \hline \end{array}\ ] ] the dissection shown in fig .
[ fig:3dpiece ] can be described algebraically as follows .
we describe it in the more logical direction , going from the triangular prism to the tetrahedron since this is what we will generalize to higher dimensions in the next section .
the input is a vector @xmath119 with @xmath120 , @xmath121 ; the output is a vector @xmath122 with @xmath123 , given by @xmath124 @xmath125 the four cases in eqn .
( [ eqf3 ] ) , after being transformed , correspond to the pieces labeled 4 , 3 , 2 , 1 respectively in fig .
[ fig:3dpiece ] .
we see from eqn .
( [ eqf3 ] ) that in the second and third cases the linear transformation has determinant @xmath50 , indicating that these two pieces must be reflected .
since it is hard to visualize dissections in dimensions greater than three , we give a schematic representation of the above dissection that avoids drawing polyhedra .
[ fig : shift ] shows a representation of the transformation from the triangular prism to the tetrahedron @xmath64 , equivalent to that given in eqn .
( [ eqf3 ] ) .
the steps shown in fig . [ fig : shift ] may be referred to as `` cut and paste '' operations , because , as fig .
[ fig : shift ] shows , the vector in the triangular prism is literally cut up into pieces which are rearranged and relabeled .
note that , to complete the transformation , we precede this operation by the dissection given in eqn .
( [ eq2db ] ) , finally establishing the bijection between @xmath107 and @xmath64 .
[ fig5 ] we now describe the mapping shown in fig . [
fig : shift ] in more detail .
the triangular prism is represented by the set of partially ordered triples @xmath126 with @xmath127 and @xmath128 , and we wish to transform this into the tetrahedron consisting of the points @xmath122 with @xmath129 .
we divide the interval @xmath130 into @xmath54 equal segments of length @xmath131 , and consider where the points @xmath132 and @xmath133 fall in this interval , given that @xmath119 is in the triangular prism .
there are three possibilities for where @xmath133 lies in relation to @xmath134 , and we further divide the case @xmath135 into two subcases depending on whether @xmath136 or @xmath137 .
these are the four cases shown in fig .
[ fig : shift ] , and correspond one - to - one with the four dissection pieces in fig .
[ fig:3dpiece ] .
[ fig : shift ] shows how the triples @xmath138 ( reindexed according to their relative positions ) are mapped to the triples @xmath139 .
the last column of fig .
[ fig : shift ] shows the ranges of the @xmath140 in the four cases ; the fact that these ranges are disjoint guarantees that the mapping from @xmath138 to @xmath139 is invertible .
the ranges of the @xmath140 will be discussed in more detail in the following section after the general algorithms are presented .
this operation can now be described without explicitly mentioning the underlying dissection .
each interval of length @xmath141 , together with the given @xmath82 values within it , is treated as a single complete unit . in the cut and paste "
operations , these units are rearranged and relabeled in such a way that the operation is invertible .
in the previous section we provided an encoding and decoding algorithm for weights @xmath53 and @xmath54 , based on our geometric interpretation of @xmath88 and @xmath142 as points in @xmath42 . in this section , the algorithm is generalized to larger values of the weight @xmath0 .
we start with the geometry , and give a dissection of the `` brick '' @xmath74 into the orthoscheme @xmath61 .
we work with the normalized coordinates @xmath143 ( for a point in @xmath74 ) and @xmath144 ( for a point in @xmath61 ) , where @xmath145 . later in this section ,
we discuss the modifications needed to take into account the fact that the @xmath59 must be integers . restating the problem ,
we wish to find a bijection @xmath146 between the sets @xmath74 and @xmath61 .
the inductive approach developed for @xmath54 ( where the @xmath53 case was a subproblem ) will be generalized .
of course the bijection @xmath147 between @xmath148 and @xmath149 is trivial .
we assume that a bijection @xmath150 is known between @xmath151 and @xmath152 , and show how to construct a bijection @xmath146 between @xmath74 and @xmath61 .
the last step in the induction uses a map @xmath153 from the prism @xmath154 to @xmath155 ( @xmath156 is the map described in eqn .
( [ eq2db ] ) and @xmath157 is described in eqn .
( [ eqf3 ] ) ) .
the mapping @xmath146 from @xmath74 to @xmath61 is then given recursively by @xmath158 , where @xmath159 for @xmath160 we set @xmath161 by iterating eqn .
( [ eqf ] ) , we see that @xmath146 is obtained by successively applying the maps @xmath162 , @xmath153 .
the following algorithm defines @xmath153 for @xmath163 .
we begin with an algebraic definition of the mapping and its inverse , and then discuss it further in the following section .
the input to the mapping @xmath153 is a vector @xmath164 , with @xmath165 and @xmath166 ; the output is a vector @xmath167 . _ forward mapping @xmath153 _ ( @xmath163 ) : \1 ) let @xmath168 \2 ) let @xmath169 \3 ) set @xmath170 equal to : @xmath171 eqn .
( [ eqn - f - shift ] ) identifies the cut and paste " operations required to obtain @xmath172 for different ranges of the variable @xmath173 . if the initial index in one of the four cases in eqn .
( [ eqn - f - shift ] ) is smaller than the final index , that case is to be skipped . a case
is also skipped if the subscript for an @xmath82 is not in the range @xmath174 . note in step 1 that @xmath175 if @xmath176 is the largest of the @xmath82 s .
this implies that @xmath177 , and then step 3 is the identity map .
the inverse mapping @xmath178 from @xmath61 to @xmath74 has a similar recursive definition .
the @xmath0th step in the induction is the map @xmath179 defined below .
for @xmath160 we set @xmath180 the map @xmath178 is obtained by successively applying the maps @xmath181 , @xmath182
. _ inverse mapping @xmath183 _ ( @xmath163 ) : \1 ) let @xmath184 \2 ) if @xmath185 , let @xmath186 , otherwise let @xmath187 in either case , let @xmath188 .
\3 ) set @xmath189 equal to : @xmath190 note that the transformations in eqn .
( [ eqn - f - shift ] ) and eqn .
( [ eqn - g - shift ] ) are formal inverses of each other , and that these transformations are volume - preserving .
the underlying linear transformations are orthogonal transformations with determinant @xmath49 or @xmath50 . before proceeding further ,
let us verify that in the case @xmath54 , the mapping @xmath191 agrees with that given in eqn .
( [ eqf3 ] ) . * if @xmath192 , then @xmath193 , @xmath177 and the map is the identity , as mentioned above . * if @xmath194 there are two subcases : * * if @xmath195 then @xmath196 , @xmath177 . * * if @xmath197 then @xmath196 , @xmath198 . *
if @xmath199 , then @xmath200 , @xmath177 . the transformations in eqn .
( [ eqn - f - shift ] ) now exactly match those in eqn .
( [ eqf3 ] ) . in fig .
[ fig : intuitive ] , we give a graphical interpretation of the algorithm , which can be regarded as a generalization of the `` cut and paste '' description given above .
this figure shows the transformation defined by the @xmath0th step @xmath153 in the algorithm . at this step ,
we begin with a list of @xmath201 numbers @xmath202 in increasing order , and a further number @xmath176 which may be anywhere in the interval @xmath203 . this list of @xmath0 numbers is plotted in the plane as the set of @xmath0 points @xmath204 for @xmath205 ( indicated by the solid black circles in fig .
[ fig : intuitive ] ) . in the first step in the forward algorithm ,
the augmented list @xmath206 is sorted into increasing order . in the sorted list
, @xmath176 now occupies position @xmath207 , so the point @xmath208 moves to the left , to the new position @xmath209 , and the points @xmath210 for @xmath211 move to the right .
this is indicated by the arrows in the figure .
the new positions of these points are marked by hollow circles .
the point @xmath209 now lies between the grid points @xmath212 and @xmath213 ( it may coincide with the latter point ) , since @xmath214 .
we draw the line @xmath215 ( shown as the dashed - and - dotted line in fig .
[ fig : intuitive ] ) .
this has unit slope and passes through the points @xmath213 and @xmath216 .
the algorithm then computes @xmath217 to be the smallest index @xmath60 for which @xmath218 is on or above this line . once @xmath219 and @xmath220 have been determined , the forward mapping proceeds as follows .
the points @xmath210 for @xmath221 are shifted to the right of the figure and are moved upwards by the amount @xmath222 , their new positions being indicated by crosses in the figure .
finally , the origin is moved to the grid point @xmath223 and the points are reindexed .
the @xmath224 points which originally had indices @xmath225 become points @xmath226 after reindexing . in the new coordinates ,
the final positions of the points lie inside the square region @xmath227 .
the reader can check that this process is exactly equivalent to the algebraic description of @xmath153 given above . to recover @xmath219 and @xmath220
, we first determine the value of @xmath224 .
this can indeed be done since @xmath228 is precisely the index of the largest @xmath229 that lies on or above the line @xmath230 in the new coordinate system .
note that the position of this line is independent of @xmath219 and @xmath220 and @xmath231 .
this works because the points @xmath232 in the original coordinate system , before the origin is shifted , are moved right by @xmath0 units and upwards by @xmath0 units , so points below the dashed - and - dotted line remain below the line .
furthermore , observe that in the new coordinate system the number of points @xmath233 below the line @xmath234 is equal to @xmath235 .
thus the correct @xmath219 and @xmath220 values may be recovered , and the inverse mapping can be successfully performed .
the following remarks record two properties of the algorithm that will be used later .
_ remark 1 : _ step 2 of the forward algorithm implies that @xmath236 and @xmath237 .
it follows that there is no @xmath60 in the range @xmath238 for which @xmath239 _ remark 2 : _ the forward algorithm produces a vector @xmath240 whose components satisfy @xmath241 @xmath242 and @xmath243 eqns .
( [ eqn : moveleft ] ) and ( [ eqn : moveright ] ) follow from the minimizations in steps 1 and 2 of the forward algorithm , respectively . the right - hand side of eqn .
( [ eqbelow ] ) expresses the fact , already mentioned , that the first @xmath220 points remain below the dotted - and - dashed line after they are shifted .
we now give the formal proof that the algorithm is correct .
this is simply a matter of collecting together facts that we have already observed .
[ th1 ] for any @xmath244 , the forward mapping @xmath153 is a one - to - one mapping from @xmath245 to @xmath61 with inverse @xmath183 .
first , it follows from remark 2 that , for @xmath246 , @xmath247 satisfies @xmath248 , and so is an element of @xmath61 .
suppose there were two different choices for @xmath33 , say @xmath249 and @xmath250 , such that @xmath251 we know that @xmath252 determines @xmath253 and @xmath219 .
so @xmath249 and @xmath250 have the same associated values of @xmath219 and @xmath220 .
but for a given pair @xmath254 , eqn .
( [ eqn - f - shift ] ) is invertible .
hence @xmath255 , and @xmath153 is one - to - one .
note that the transformations in eqn .
( [ eqn - f - shift ] ) and eqn .
( [ eqn - g - shift ] ) are inverses of each other . hence @xmath153 is also an onto map , and @xmath183 is its inverse .
the map @xmath153 , which dissects the prism @xmath154 to give the orthoscheme @xmath155 , has one piece for each pair @xmath256 . if @xmath257 then @xmath177 , while if @xmath258 , @xmath220 takes all values from @xmath259 to @xmath260 .
( it is easy to write down an explicit point in the interior of the piece corresponding to a specified pair of values of @xmath219 and @xmath220 .
assume @xmath261 and set @xmath262 .
take the point with coordinates @xmath263 given by @xmath264 ; @xmath265 for @xmath266 ; @xmath267 for @xmath268 ; @xmath269 for @xmath270 . )
the total number of pieces in the dissection is therefore @xmath271 which is @xmath272 for @xmath273 .
this is a well - known sequence , entry a124 in @xcite , which by coincidence also arises in a different dissection problem : it is the maximal number of pieces into which a circular disk can be cut with @xmath201 straight cuts .
for example , with three cuts , a pizza can be cut into a maximum of seven pieces , and this is also the number of pieces in the dissection defined by @xmath274 . to apply the above algorithm to the problem of encoding and decoding constant weight codes , we must work with positive integers rather than real numbers , which entails a certain loss in rate , although the algorithms remain largely unchanged .
let @xmath275 , and let @xmath1 and @xmath0 be given with @xmath276 . in a manner analogous to the real - valued case , we find a bijection between a finite hyper - rectangle or brick @xmath277 and a subset of the finite orthoscheme @xmath278 , where @xmath279 is the set of vectors @xmath280 satisfying @xmath281 for @xmath205 , and @xmath282 is the set of vectors @xmath280 satisfying @xmath283 note that usually @xmath284 , which entails a loss in rate .
the forward mapping @xmath153 is now replaced by the map @xmath285 , which sends @xmath286 with @xmath287 and @xmath288 to an element of @xmath289 .
let us write @xmath290 , where @xmath291 and @xmath292 .
we partition the range @xmath293 into @xmath0 parts , where the first @xmath294 parts each have @xmath295 elements , the next @xmath296 parts each have @xmath297 elements , and the last part has @xmath295 elements ( giving a total of @xmath1 elements ) .
this is similar to the real - valued case , where each interval had length @xmath141 .
\1 ) let @xmath298 \2 ) let @xmath299 where @xmath300 .
\3 ) set @xmath301 equal to : @xmath302 the inverse mapping @xmath183 is similarly replaced by the map @xmath303 @xmath304 , defined as follows . again , assume @xmath290 .
\1 ) let @xmath305 where @xmath306 .
\2 ) if @xmath185 , let @xmath186 , otherwise let @xmath307 in either case , let @xmath188 .
\3 ) set @xmath308 equal to : @xmath309 we omit the proofs , since they are similar to those for the real - valued case .
the overall complexity of the transform algorithm is @xmath2 , because at each induction step the complexity is linear in the weight at that step .
recall that the complexities of the arithmetic coding method and knuth s complementation method are both @xmath32 .
thus when the weight @xmath0 is larger than @xmath310 , the geometric approach is less competitive .
when the weight is low , the proposed geometric technique is more efficient , because knuth s complementation method is not applicable , while the dissection operations of the proposed algorithm makes it faster than the arithmetic coding method .
furthermore , due to the structure of the algorithm , it is possible to parallelize part of the computation within each induction step to further reduce the computation time .
so far little has been said about mapping a binary sequence to an integer sequence @xmath311 such that @xmath312 $ ] , where @xmath313 and @xmath314 are the lower and upper bound of the valid range as specified by the algorithm .
a straightforward method is to treat the binary sequence as an integer number and then use quotient and remainder " method to find such a mapping .
however , this requires a division operation , and when the binary sequence is long , the computation is not very efficient .
a simplification is to partition the binary sequence into short sequences , and map each short binary sequence to a pair of integers , as in the case of a weight two constant weight codes . through proper pairing of the ranges ,
the loss in the rate can be minimized .
the overall rate loss has two components , the first from the rounding involved in using natural numbers , the second from the loss in the above simplified translation step .
however , when the weight is on the order of @xmath310 , and @xmath1 is in the range of @xmath315 , the rate loss is usually @xmath316 bits per block .
for example , when @xmath317 , @xmath318 , then the rate loss is 2 bits / block compared to the best possible code which would encode @xmath319 information bits .
we propose a novel algorithm for encoding and decoding constant weight binary codes , based on dissecting the polytope defined by the set of all binary words of length @xmath1 and weight @xmath0 , and reassembling the pieces to form a hyper - rectangle corresponding to the input data .
the algorithm has a natural recursive structure , which enables us to give an inductive proof of its correctness .
the proposed algorithm has complexity @xmath2 , independent of the length of the codewords @xmath1 .
it is especially suitable for constant weight codes of low weight . | we present a novel technique for encoding and decoding constant weight binary codes that uses a geometric interpretation of the codebook .
our technique is based on embedding the codebook in a euclidean space of dimension equal to the weight of the code .
the encoder and decoder mappings are then interpreted as a bijection between a certain hyper - rectangle and a polytope in this euclidean space .
an inductive dissection algorithm is developed for constructing such a bijection .
we prove that the algorithm is correct and then analyze its complexity .
the complexity depends on the weight of the code , rather than on the block length as in other algorithms .
this approach is advantageous when the weight is smaller than the square root of the block length .
constant weight codes , encoding algorithms , dissections , polyhedral dissections , bijections , mappings , dehn invariant . |
it is well - known that various observables associated with inclusive rare @xmath19-decays like @xmath20 and @xmath21 sensitively depend on potential new physics contributions .
but even in the absence of new physics these observables are important , because they provide checks on the one - loop structure of the standard model ( sm ) theory and can be used to gain information on the cabibbo - kobayashi - maskawa ( ckm ) matrix elements @xmath22 and @xmath23 , which are difficult to measure directly . at present , a lot of data already exists on @xmath24 @xcite and on @xmath25 @xcite and it is expected that in the future also data on the ckm suppressed counterparts , i.e. on @xmath26 and on @xmath27 will become available .
the same holds for experimental information on additional observables , like cp - rate asymmetries or forward - backward asymmetries in the decays @xmath21 . in order to fully exploit and interpret the experimental data , it is obvious that precise calculations in the sm ( or certain extensions thereof ) are needed . the main problem in the theoretical description of the decay @xmath28 is due to the long - distance contributions induced by @xmath29 resonant states and in principle also by @xmath30 resonant states
the latter are , however , strongly ckm suppressed .
this suppression is not present in the case of @xmath7 , as the ckm factors involved in the contributions from @xmath29 and @xmath30 resonant states are of the same order . when the invariant mass @xmath31 of the lepton pair is close to the mass of a resonance , only model dependent predictions for such long distance contributions are available at present .
it is therefore unclear whether the theoretical uncertainty can be reduced to less than @xmath32 when integrating over these domains @xcite .
however , restricting @xmath31 to a region below the @xmath29 resonances , the long distance effects in @xmath3 are under control .
the same is true for @xmath7 when choosing a region of @xmath31 which is below the @xmath14- and above the @xmath33-resonance regions .
it turns out that in those ranges of @xmath31 the corrections to the pure perturbative picture can be analyzed within the heavy quark effective theory ( hqet ) .
in particular , all available studies indicate that for the region @xmath34 the non - perturbative effects are below 10@xmath35 @xcite .
consequently , observables like differential decay rates , forward - backward asymmetries and cp - rate asymmetries for @xmath36 can be precisely predicted in this region of @xmath31 using renormalization group improved perturbation theory .
it was pointed out in the literature that the differential decay rate and the forward - backward asymmetry in @xmath3 are particularly sensitive to new physics in this kinematical window @xcite . in the context of the sm there
exist computations of next - to - leading logarithmic ( nll ) qcd corrections to the branching ratios for @xmath37 @xcite and @xmath38 and the corresponding cp - rate asymmetries @xcite .
next - to - next - to - leading logarithmic ( nnll ) qcd corrections to the branching ratio @xcite and the forward - backward asymmetry in @xmath39 are also available @xcite .
for a recent review see e.g. @xcite . the corresponding nnll results for the process @xmath7 are missing , however .
the aim of the present paper is to close this gap .
the main difference between the calculations for @xmath3 and @xmath40 lies in the contributions of the current - current operators . in the existing nnll calculations of @xmath3
only those associated with @xmath41 and @xmath42 were included at the two - loop level because those induced by @xmath4 and @xmath5 are strongly ckm suppressed ( see section [ section : eff ] for the definition of the operators @xmath43 ) .
for @xmath7 the contributions generated by @xmath4 and @xmath5 are no longer ckm suppressed and have to be taken into account as well . at first sight
, it seems that the two - loop matrix elements of @xmath4 and @xmath5 can be straightforwardly obtained from those of @xmath41 and @xmath42 by simply taking the limit @xmath44 .
this is , however , not possible for some of the diagrams in fig .
[ fig : o1o2 ] , because the two - loop matrix elements of @xmath41 and @xmath42 were derived by doing various expansions .
in particular , one of the expansion parameters is @xmath45 , which is formally @xmath46 when restricting @xmath31 to the window discussed above .
obviously , the analogous quantity for the @xmath6-quark contribution , @xmath47 , can not be used as an expansion parameter , which implies that genuinely new calculations for the @xmath6-quark contributions are needed . as discussed in section [ section : o1o2 ] , the calculations of certain diagrams associated with @xmath48 are even more involved than those associated with @xmath49 . to derive the new results , we used dimension - shifting techniques in order to reduce certain tensor integrals to scalar ones and integration - by - parts techniques to further simplify the scalar integrals @xcite . as the main emphasis of this paper is the derivation of the matrix elements @xmath50 at order @xmath51 , we keep the phenomenological analysis relatively short . in particular , we do not take into account power corrections , but merely illustrate how the nnll contributions modify the scale dependences of the branching ratio , the forward - backward asymmetry and the cp - rate asymmetry .
a more detailed phenomenology , including power corrections , will be presented elsewhere .
the paper is organized as follows : in section [ section : eff ] we present the effective hamiltonian for the decay @xmath52 .
section [ section : o1o2 ] is devoted to the virtual @xmath53 corrections to the operators @xmath54 and @xmath55 .
subsequently , section [ section : o789 ] presents the corresponding contributions to the form factors of the operators @xmath56 and @xmath57 . with these results at hand ,
we discuss in section [ section : decay_width ] the corrections to the decay width of @xmath58 . in section [ section
: phenom ] we show some applications of our results .
a summary of the paper is presented in section [ section : summary ] .
the appendices contain technical details about the performed calculation : appendix [ section : techniques ] explains the dimension - shifting and integration - by - parts techniques .
these techniques are then applied to the calculation of diagrams [ fig : o1o2]d ) , which is presented in appendix [ section : diagram_d ] .
appendix [ appendix : wilson ] outlines a procedure on how to calculate the evolution matrix for the wilson coefficients as a power series in @xmath51 .
appendix [ appendix : oneloop ] contains one - loop matrix elements needed in the calculation of the counterterms .
finally , in appendix [ appendix : brems ] we present the results for those bremsstrahlung contributions which are free of infrared and collinear divergences .
the appropriate framework for studying qcd corrections to rare @xmath19-decays in a systematic way is the effective hamiltonian technique . for the specific decay channels @xmath59 and @xmath7 ( @xmath60 , @xmath61 ) ,
the effective hamiltonian is derived by integrating out the @xmath62-quark , the @xmath63-boson and the @xmath64-boson . in the process @xmath65
, the appearing ckm combinations are @xmath66 , @xmath67 and @xmath68 , where @xmath69 .
since @xmath70 is much smaller than @xmath71 and @xmath72 , it is a safe approximation to set @xmath66 equal to zero . using then the unitarity properties of the ckm matrix
, the ckm dependence of the hamiltonian can be written as a global factor @xmath68 . in the case of @xmath7 ,
all three quantities @xmath73 ( @xmath74 ) are of the same order of magnitude .
therefore , as no approximation is possible , the ckm dependence does not globally factorize .
the effective hamiltonian reads @xmath75 \ , .\ ] ] we choose the operator basis according to @xcite : @xmath76 where the subscripts @xmath77 and @xmath78 refer to left- and right - handed components of the fermion fields , respectively .
the factors @xmath79 in the definition of the operators @xmath80 , @xmath81 and @xmath57 as well as the factor @xmath82 present in @xmath83 have been chosen by misiak @xcite in order to simplify the organization of the calculation . with these definitions ,
the one - loop anomalous dimensions [ needed for a leading logarithmic ( ll ) calculation ] of the operators @xmath84 are all proportional to @xmath85 , while two - loop anomalous dimensions [ needed for a next - to - leading logarithmic ( nll ) calculation ] are proportional to @xmath86 , etc .
after this important remark we now outline the principal steps which lead to a ll , nll , and a nnll prediction for the decay amplitude for @xmath87 : 1 .
a matching calculation between the full sm theory and the effective theory has to be performed in order to determine the wilson coefficients @xmath88 at the high scale @xmath89 . at this scale ,
the coefficients can be worked out in fixed order perturbation theory , i.e. they can be expanded in @xmath85 : @xmath90 at ll order , only @xmath91 are needed , at nll order also @xmath92 , etc .
the coefficient @xmath93 was worked out in refs .
@xcite , while @xmath94 and @xmath95 were calculated in ref .
2 . the renormalization group equation ( rge ) has to be solved in order to get the wilson coefficients at the low scale @xmath96 . for this rge step the anomalous dimension matrix @xmath97 , which can be expanded as @xmath98 is required up to the term proportional to @xmath99 when aiming at a nnll calculation .
after the matching step and the rge evolution , the wilson coefficients @xmath100 can be decomposed into a ll , nll and nnll part according to @xmath101 we stress at this point that the entries in @xmath99 which describe the three - loop mixings of the four - quark operators @xmath102 into the operator @xmath81 have been calculated only recently @xcite . in order to include the impact of these new ingredients on the wilson coefficient @xmath103
, we had to reanalyze the rge step . in appendix
[ appendix : wilson ] , we derive a practical formula for the evolution matrix @xmath104 at nnll order , generalizing existing formulas at nll order ( see e.g. @xcite ) .
3 . in order to get the decay amplitude
, the matrix elements @xmath105 have to be calculated . at ll precision , only the operator @xmath81 contributes , as this operator is the only one which at the same time has a wilson coefficient starting at lowest order and an explicit @xmath79 factor in the definition .
hence , at nll precision , qcd corrections ( virtual and bremsstrahlung ) to the matrix element of @xmath81 are needed .
they have been calculated in refs .
@xcite . at nll precision ,
also the other operators start contributing , viz . @xmath106 and
@xmath107 contribute at tree - level and the four - quark operators @xmath108 at one - loop level .
accordingly , qcd corrections to the latter matrix elements are needed for a nnll prediction of the decay amplitude .
the formally leading term @xmath109 to the amplitude for @xmath87 is smaller than the nll term @xmath110 \ , c_9^{(1)}(\mu_b)$ ] @xcite .
we adapt our systematics to the numerical situation and treat the sum of these two terms as a nll contribution .
this is , admittedly some abuse of language , because the decay amplitude then starts out with a term which is called nll . as pointed out in step 3 ) , @xmath8 qcd corrections to the matrix elements @xmath111 have to be calculated in order to obtain the nnll prediction for the decay amplitude . in the present paper
we systematically evaluate virtual corrections of order @xmath51 to the matrix elements of @xmath112 , @xmath113 , @xmath80 , @xmath83 , @xmath81 and @xmath57 .
as the wilson coefficients of the gluonic penguin operators @xmath114 are much smaller than those of @xmath112 and @xmath113 , we neglect qcd corrections to their matrix elements .
we also systematically include gluon bremsstrahlung corrections to the matrix elements of the operators just mentioned .
some of these contributions contain infrared and collinear singularities , which are canceled when combined with the virtual corrections .
associated with the operators @xmath54 and @xmath55 . the fermions ( @xmath117- , @xmath118- , @xmath6- and @xmath119-quarks )
are represented by solid lines , whereas the curly lines represent gluons .
the circle - crosses denote the possible locations where the virtual photon ( which then splits into a lepton pair ) is emitted . ] in this section we present the calculation of the virtual @xmath0 corrections to the matrix elements of the current - current operators @xmath54 and @xmath55 . using the naive dimensional regularization scheme ( ndr ) in @xmath120 dimensions ,
both ultraviolet and infrared singularities show up as @xmath121 poles ( @xmath122 ) .
the ultraviolet singularities cancel after including the counterterms .
collinear singularities are regularized by retaining a finite down quark mass @xmath123 .
they are canceled together with the infrared singularities at the level of the decay width when taking the bremsstrahlung process @xmath124 into account .
we use the @xmath125 renormalization scheme , i.e. we introduce the renormalization scale in the form @xmath126 followed by minimal subtraction . the precise definition of the evanescent operators , which is necessary to fully specify the renormalization scheme , will be given later .
gauge invariance implies that the qcd corrected matrix elements of the operators @xmath127 can be written as @xmath128 where @xmath129 and @xmath130 are the tree - level matrix elements of @xmath81 and @xmath80 , respectively .
equivalently , we may write @xmath131,\ ] ] where the operators @xmath132 and @xmath133 are defined as @xmath134 we present the final results for the qcd corrected matrix elements in the form of eq .
( [ formdeftilde ] ) .
the full set of the diagrams contributing at @xmath0 to the matrix elements @xmath135 is shown in fig .
[ fig : o1o2 ] .
as indicated , the diagrams associated with @xmath54 and @xmath55 are topologically identical .
they differ only in the color structure .
while the matrix elements of the operator @xmath55 all involve the color structure @xmath136 there are two possible color structures for the corresponding diagrams of @xmath54 , viz @xmath137 the structure @xmath138 appears in diagrams [ fig : o1o2]a)-d ) , and @xmath139 enters diagrams [ fig : o1o2]e ) and [ fig : o1o2]f ) . using the relation @xmath140 we find that @xmath141 and @xmath142 , with @xmath143 inserting @xmath144 , the color factors are @xmath145 and @xmath146 .
the contributions from @xmath54 are obtained by multiplying those from @xmath55 by the appropriate factors , i.e. by @xmath147 and @xmath148 , respectively . as the renormalized @xmath0 contributions of the operators @xmath41 and @xmath42 are discussed in detail in ref .
@xcite , we only discuss the calculations of the contributions from @xmath4 and @xmath5 to the individual form factors .
the rest of this section is organized as follows : we discuss the calculations of the diagrams [ fig : o1o2]a)-e ) for the operators @xmath48 .
notice that all results are given as an expansion in the small quantity @xmath149 , where @xmath11 is the invariant mass squared of the lepton pair , and that we keep only terms up to @xmath150 .
after deriving the counterterms that cancel the divergences of the diagrams mentioned above , we present the renormalized contributions to the form factors . we postpone the discussion of diagrams [ fig : o1o2]f ) as it turns out to be more convenient to take them into account when discussing the virtual corrections to @xmath81 .
the calculation of the contributions to @xmath151 and @xmath152 from the diagrams in figs .
[ fig : o1o2]a ) and [ fig : o1o2]b ) opposes no difficulties , as it can be performed by using the mellin - barnes approach @xcite .
alternatively , one may get the results directly from the corresponding form factors of the @xmath153 transition by taking the limit @xmath154 .
the form factors associated with the diagrams in fig . [ fig : o1o2]a ) are given by @xmath155 = c_f \cdot \bigg [ & -\frac{2}{27\,\epsilon^2 } + \left(\frac{1}{\epsilon } + 4\,l_\mu \right ) \left(-\frac{19}{81 } + \frac{4}{27}\,l_s - \frac{4}{27}\,i\pi \right ) - \frac{8}{27\,\epsilon}\,l_\mu { \nonumber}\\ & - \frac{16}{27}\,l_\mu^2 + \left ( -\frac{463}{486 } - \frac{38\,i\pi}{81 } + \frac{5\,\pi^2}{27 } \right ) - \frac{4}{27}\,{\hat{s}}+ \left ( -\frac{1}{27 } - \frac{2}{27}\ ,
l_s \right ) { \hat{s}}^2 { \nonumber}\\ & + \left(-\frac{4}{243}-\frac{8}{81}\ , l_s \right ) { \hat{s}}^3 + \frac{26}{81}\ , l_s + \frac{8}{27}\ , i \pi\,l_s - \frac{2}{27}\ , l_s^2 \bigg ] , \\ { \nonumber}\\ f_{2,u}^{(7)}[a ]
= c_f \cdot \bigg [ & \frac{1}{27}\left(\frac{1}{\epsilon } + 4\,l_\mu \right ) + \frac{37}{162 } + \frac{2}{27}\,i\pi + \frac{2}{27}\,{\hat{s}}\left ( 1+{\hat{s}}+{\hat{s}}^2 \right ) l_s \bigg],{\nonumber}\end{aligned}\ ] ] where @xmath156 for the sum of the diagrams in fig .
[ fig : o1o2]b ) we find @xmath157 = c_f \cdot \bigg [ & -\frac{2}{27\,\epsilon^2 } + \left(\frac{1}{\epsilon } + 4\,l_\mu \right ) \left ( \frac{1}{81 } - \frac{4}{135}\,{\hat{s}}- \frac{1}{315}\,{\hat{s}}^2 - \frac{4}{8505}\,{\hat{s}}^3\right ) - \frac{8}{27\,\epsilon}\ , l_\mu { \nonumber}\\ & - \frac{16}{27}\ , l_\mu^2 + \left(\frac{917}{486 } - \frac{19\,\pi^2}{81 } \right ) + \left ( \frac{172}{225 } - \frac{2\,\pi^2}{27 } \right){\hat{s}}{\nonumber}\\ & + \left ( -\frac{871057}{396900 } + \frac{2\,\pi^2}{9 } \right ) { \hat{s}}^2 + \left ( -\frac{83573783}{10716300 } + \frac{64\,\pi^2}{81 } \right ) { \hat{s}}^3 \bigg],\end{aligned}\ ] ] @xmath158 = c_f \cdot \bigg [ & -\frac{5}{27\,\epsilon } - \frac{20}{27}\ , l_\mu { \nonumber}\\ & + \frac{13}{162 } + \left ( \frac{25}{81 } - \frac{\pi^2}{27 } \right ) { \hat{s}}+ \left ( \frac{118}{81 } - \frac{4\,\pi^2}{27 } \right ) { \hat{s}}^2 + \left ( \frac{10361}{2835 } - \frac{10\,\pi^2}{27 } \right){\hat{s}}^3 \bigg ] . { \nonumber}\end{aligned}\ ] ] the computation of the diagrams in fig . [ fig : o1o2]d ) is by far the most complicated piece in our entire calculation of the @xmath159 corrections to the matrix element for @xmath160 . after various unsuccessful attempts
, we managed to obtain the result by using the dimension - shifting method @xcite ( see appendix [ section : tensor ] ) , combined with the method of partial integration ( see appendix [ section : intbyparts ] ) .
since we want to include the details of the actual calculation , we relegate them to appendix [ section : diagram_d ] . here
, we merely present the final results , viz the contributions to the form factors , which read @xmath161 = c_f \cdot \bigg [ & \frac{2}{3\,\epsilon^2 } + \frac{1}{\epsilon } \left ( \frac{5}{3 } - \frac{4\,l_s}{3 } + \frac{8}{3}\,l_\mu + \frac{4\,i\pi}{3 } \right ) + \frac{16}{3}\,l_\mu^2 + \frac{7}{6 } -4\,l_s \\ & + \frac{2}{3}\,l_s^2 + 4\,i\pi - \frac{4\,i\pi}{3 } \ , l_s - \pi^2 + \left(\frac{20}{3 } - \frac{16}{3}\,l_s + \frac{16\,i\pi}{3 } \right ) \ , l_\m { \nonumber}\\ & + \left ( \frac{2}{3 } + \frac{2}{3}\,l_s - \frac{2}{3}\,l_s^2 - \frac{2\,i\pi}{3 } + \frac{4\,i\pi}{3}\,l_s \right)\,{\hat{s}}+ \left ( \frac{2}{3 } + 2\,l_s -2\,i\pi\right ) \ , { \hat{s}}^2 { \nonumber}\\ & + \left(\frac{2}{3 } + \frac{10}{3}\ , l_s + \frac{4}{3 } \ , l_s^2 - \frac{10\,i\pi}{3 } - \frac{8\,i\pi}{3}\,l_s \right ) \ , { \hat{s}}^3 \bigg ] , { \nonumber}\\ { \nonumber}\\ f_{2,u}^{(7)}[d ] = c_f \cdot \bigg [ & \frac{2}{3\,\e } + \frac{7}{3 } + \frac{8}{3}\ , l_\mu - \left(\frac{1}{3 } - \frac{1}{3}\ , l_s - \frac{1}{3}\ , l_s^2 + \frac{i\pi}{3 } + \frac{2\,i\pi}{3}\ , l_s \right)\,{\hat{s}}\\ & -\left(\frac{1}{3 } + \frac{1}{3}\,l_s - \frac{1}{3}\ , l_s^2 -\frac{i\pi}{3 } + \frac{2\,i\pi}{3}\ , l_s\right)\ , { \hat{s}}^2 - \left(\frac{1}{3 } + l_s - i\pi \right)\ , { \hat{s}}^3 \bigg ] { \nonumber}.\end{aligned}\ ] ] the calculation of this diagram can be done in a very simple and efficient way .
we add the two subdiagrams and integrate out the loop momentum of the virtual gluon .
next we integrate over the remaining loop momentum , being left with a four dimensional feynman parameter integral . after introducing a single mellin - barnes representation of the occurring denominator
, the parameter integrals can all be performed . at this level ,
the result contains euler beta - functions involving the mellin - barnes parameter .
finally , the mellin - barnes integral can be resolved applying the residue theorem , which naturally leads to an expansion in the parameter @xmath162 . the contribution of the diagrams in fig .
[ fig : o1o2]c ) to the form factors reads @xmath163
= c_f \cdot \bigg [ & \frac{2}{3\,\epsilon^2 } + \frac{1}{\epsilon } \left ( \frac{5}{3 } - \frac{4\,l_s}{3 } + \frac{8}{3}\,l_\mu + \frac{4\,i\pi}{3 } \right ) + \frac{16}{3}\,l_\mu^2 \\ & + \frac{1}{2 } - 6\,l_s + \frac{2}{3}\,l_s^2 + \frac{10\,i\pi}{3 } - \frac{8\,i\pi}{3}\,l_s - \frac{5\,\pi^2}{3 } { \nonumber}\\ & + \left ( \frac{4}{3 } - 4\,l_s + \frac{2}{3}\,l_s^2 + \frac{2\,\pi^2}{9 } \right ) { \hat{s}}+ \left ( -1 - 2\,l_s + \frac{2}{3}\,l_s^2 + \frac{2\,\pi^2}{9 } \right ) { \hat{s}}^2 { \nonumber}\\ & + \left ( -\frac{41}{27 } - \frac{10}{9}\,l_s + \frac{2}{3}\,l_s^2 + \frac{2\,\pi^2}{9 } \right){\hat{s}}^3 + \left ( \frac{20}{3 } + \frac{16\,i\pi}{3 } - \frac{16}{3}\,l_s \right ) l_\mu \bigg ] , { \nonumber}\\ { \nonumber}\\ f_{2,u}^{(7)}[c ] = c_f \cdot \bigg [ & \frac{1}{3\,\epsilon } + \frac{5}{2 } + \frac{2\,i\pi}{3 } + \left ( \frac{2\,l_s}{3 } - \frac{l_s^2}{3 } - \frac{\pi^2}{9}\right ) { \hat{s}}+ \left ( \frac{2}{3 } - \frac{l_s^2}{3}-\frac{\pi^2}{9 } \right ) { \hat{s}}^2 \\ & + \left ( \frac{5}{6 } - \frac{l_s}{3 } - \frac{l_s^2}{3 } - \frac{\pi^2}{9 } \right ) { \hat{s}}^3 + \frac{4}{3}\,l_\mu \bigg ] . { \nonumber}\end{aligned}\ ] ]
we also performed the calculation of this diagram in two different , more complicated ways , namely by * using the building block @xmath164 given in @xcite and then introducing a double mellin - barnes representation , * using the dimension - shifting and integration - by - parts techniques as explained when discussing the computation of the diagrams in fig .
[ fig : o1o2]d ) ( see also appendices [ section : tensor ] and [ section : intbyparts ] ) .
we found that all three calculations yield the same result and thus serve as an excellent check for the dimension - shifting approach and for the very complicated double mellin - barnes calculation . the diagrams in fig .
[ fig : o1o2]e ) may again be solved in two ways .
the first way is to use the large external momentum expansion technique @xcite .
the second possibility is to apply the dimension - shifting and integration - by - parts procedure @xcite also for this diagram .
we do without presenting the calculation and merely give the results for the contributions to the form factors .
@xmath165 & = c_f \cdot \bigg [ \ , -\frac{2}{3 } \left(\frac{1}{\epsilon } + 4\,l_\mu \right ) - \frac{49}{9 } - \frac{4\,i\pi}{3 } + \frac{4}{3}\,l_s + \frac{16}{3 } \ , \zeta(3 ) \bigg ] , \\
f_{2,u}^{(7)}[e ] & = \ , 0 .
{ \nonumber}\end{aligned}\ ] ] so far , we have calculated the two - loop matrix elements @xmath167 ( @xmath168 ; @xmath169 ) . as the operators mix under renormalization , there are additional contributions proportional to @xmath88 .
these counterterms arise from the matrix elements of the operators @xmath170 where the operators @xmath112@xmath171 are given in eq .
( [ oper ] ) . @xmath172 and @xmath173 are evanescent operators , i.e. operators which vanish in @xmath174 dimensions . in principle
, there is some freedom in the choice of the evanescent operators .
however , as we want to combine our matrix elements with the wilson coefficients calculated by bobeth et al .
@xcite , we have to use the same definitions : @xmath175 the operator renormalization constants @xmath176 are of the form @xmath177 the coefficients @xmath178 needed for our calculation we take from refs .
@xcite and list them for @xmath168 and @xmath179 : @xmath180 we denote the counterterm contributions to @xmath160 which are due to the mixing of @xmath4 or @xmath5 into four - quark operators by @xmath181 and @xmath182 .
they can be extracted from the equation @xmath183,\ ] ] where @xmath184 runs over the four - quark operators .
the operators @xmath185 are understood to be identified with @xmath186 for @xmath187 . as certain entries of @xmath188
are zero , only the one - loop matrix elements of @xmath54 , @xmath55 , @xmath189 , @xmath172 and @xmath173 are needed . in order to keep the presentation transparent , we relegate their explicit form to appendix [ appendix : oneloop ] .
we do not repeat the renormalization of the @xmath41 and @xmath42 contributions at this place and refer to @xcite .
there is a counterterm related to the two - loop mixing of @xmath190 ( @xmath168 ) into @xmath80 , followed by taking the tree - level matrix element @xmath191 . denoting the corresponding contribution to the counterterm form factors by @xmath192 and @xmath193
, we obtain @xmath194 the counterterms which are related to the mixing of @xmath190 ( @xmath168 ) into @xmath81 can be split into two classes : the first class consists of the one - loop mixing @xmath195 , followed by taking the one - loop corrected matrix element of @xmath81 .
it is obvious that this class contributes to the renormalization of diagram [ fig : o1o2]f ) , which we take into account when discussing the virtual corrections to @xmath81 .
we proceed in the same way with the corresponding counterterm .
the second class of counterterm contributions due to @xmath195 mixing is generated by two - loop mixing of @xmath5 into @xmath81 as well as by one - loop mixing and one - loop renormalization of the @xmath196 factor in the definition of the operator @xmath81 .
we denote the corresponding contribution to the counterterm form factors by @xmath197 and @xmath198 .
we obtain @xmath199 where we used the renormalization constant @xmath200 given by @xmath201 the total counterterms @xmath202 ( @xmath168 ; @xmath203 ) , which renormalize diagrams [ fig : o1o2]a)[fig : o1o2]e ) , are given by @xmath204 explicitly they read @xmath205 \\ { \nonumber}\\ & + \frac{8}{25515 } \left[-420 - 21420\,i\pi + 252\,{\hat{s}}+ 27\,{\hat{s}}^2 + 4\,{\hat{s}}^3 \right ] l_{\mu } \\ { \nonumber}\\ { \nonumber } & -\frac{136}{81}\ , l_s^2 + \left [ \frac{16}{243 } ( -2 + 51\,i\pi ) + \frac{544}{81}\,l_\mu\right ] l_s -\frac{512}{81}\,l_\mu^2 , \\ { \nonumber}\\ { \nonumber}f_{2,u}^{\rm{ct}(7 ) } = & -f_{2,u , \,\text{div}}^{(7 ) } + \frac{2}{2835 } \left(840\ , l_{\mu } + 70\,{\hat{s}}+ 7\,{\hat{s}}^2 + { \hat{s}}^3 \right),\end{aligned}\ ] ]
@xmath206 { \nonumber}\\ { \nonumber}\\ & + \frac{4}{76545 } \left[68460 + 21420\,i\pi -252{\hat{s}}-27 { \hat{s}}^2 -4{\hat{s}}^3\right]\,l_\mu \\ { \nonumber}\\ & + \frac{68}{243}\,l_s^2 - \left [ \frac{8}{729 } ( 160 + 51\,i\pi ) + \frac{272}{243}\,l_\mu\right ] l_s + \frac{256}{243}\,l_\mu^2 , { \nonumber}\\ { \nonumber}\\ { \nonumber}f_{1,u}^{\rm{ct}(7 ) } = & -f_{1,u , \,\text{div}}^{(7 ) } - \frac{1}{8505 } \left(840\ , l_{\mu } + 70\ , { \hat{s}}+ 7\,{\hat{s}}^2 + { \hat{s}}^3 \right).\end{aligned}\ ] ] the quantities @xmath207 @xmath208 compensate the divergent parts of the form factors associated with the virtual corrections to @xmath48 .
they are given by @xmath209 + \frac{16}{81\,\epsilon}(32\,l_\mu - 17\,l_s ) , { \nonumber}\\ { \nonumber}\\ f_{2,u , \,\text{div}}^{(7 ) } & = \frac{92}{81\,\epsilon } \ , , { \nonumber}\\ \\ f_{1,u , \,\text{div}}^{(9 ) } & = -\frac{64}{243\,\epsilon^2 } - \frac{2}{76545\,\epsilon } \big [ 71820 + 21420\,i\pi - 252\,{\hat{s}}- 27\,{\hat{s}}^2 - 4\,{\hat{s}}^3 \big ] - \frac{8}{243\,\epsilon}(32\,l_\mu - 17\,l_s ) , { \nonumber}\\ { \nonumber}\\ f_{1,u , \,\text{div}}^{(7 ) } & = -\frac{46}{243\,\epsilon } \ , .
{ \nonumber}\end{aligned}\ ] ] as mentioned before , we will take diagram [ fig : o1o2]f ) into account only in section [ section : o789 ] .
the same holds for the counterterms associated with the @xmath117- and @xmath118-quark wave function renormalization and , as stated earlier in this subsection , the @xmath0 correction to the matrix element of @xmath210 .
the sum of these contributions is @xmath211 , \quad \delta \bar z_\psi = \sqrt { z_\psi(m_b ) z_\psi(m_d ) } - 1,\ ] ] and provides the counterterm that renormalizes diagram [ fig : o1o2]f ) .
we use on - shell renormalization for the external @xmath117- and @xmath118-quark . in this scheme
the field strength renormalization constants are given by @xmath212 so far , we have discussed the counterterms which renormalize the @xmath0 corrected matrix elements @xmath213 ( @xmath168 ) . the corresponding one - loop matrix elements [ of @xmath214
are renormalized by adding the counterterms @xmath215 we now have all ingredients necessary to present the renormalized form factors associated with the operators @xmath4 and @xmath5 .
we stress again that only the contributions of the diagrams [ fig : o1o2]a)-e ) and the counterterms discussed in subsection [ countertermso2 ] are accounted for in the result below .
diagram [ fig : o1o2]f ) and associated counterterms will be included in the discussion of the virtual corrections to @xmath81 .
we decompose the renormalized matrix elements of @xmath84 ( @xmath168 ) as @xmath218,\ ] ] where the operators @xmath132 and @xmath133 are defined in eq .
( [ otilde ] ) .
the renormalized form factors read : @xmath219 with @xmath156 as has been mentioned before , we only include terms up to @xmath150 in the result .
we checked , however , that the terms of order @xmath220 are numerically negligible .
the virtual corrections to the matrix elements of @xmath80 , @xmath83 , @xmath81 and @xmath57 and their renormalization are discussed in detail in refs . @xcite . for completeness
we list the results of the renormalized matrix elements .
they may all be decomposed according to @xmath225,\ ] ] where @xmath226 { \widetilde{c}}_7^{(0 ) } = & \ , c_7^{(1 ) } , \quad { \widetilde{c}}_8^{(0 ) } = c_8^{(1 ) } , { \nonumber}\\[0.5ex ] { \widetilde{c}}_9^{(0 ) } = & \ , \frac{4\,\pi}{\alpha_s } \left(c_9^{(0 ) } + \frac{\alpha_s}{4\,\pi } c_9^{(1)}\right ) \quad\text{and}\quad { \widetilde{c}}_{10}^{(0 ) } = c_{10}^{(1)}. { \nonumber}\end{aligned}\ ] ] the renormalized corrections to the form factors @xmath227 and @xmath228 are given by @xmath229 the function @xmath230 collects the infrared- and collinear singular parts : @xmath231^{2\epsilon}}}{\epsilon_{{\text{ir } } } } \ , \frac{8}{3 } \left(1 + { \hat{s}}+ \frac{1}{2}\ , { \hat{s}}^2 + \frac{1}{3}\ , { \hat{s}}^3\right ) + \frac{{\left [ \frac{\mu}{m_b } \right]^{2\epsilon}}}{\epsilon_{{\text{ir } } } } \ , \frac{4}{3 } \ , \ln(r ) + \frac{2}{3 } \ln(r ) - \frac{2}{3 } \ln^2(r),\ ] ] where @xmath232 and @xmath233 regularize the infrared- and collinear singularities , respectively .
the renormalized corrections to the form factors of the matrix element of @xmath83 are @xmath234 @xmath235 the renormalized matrix elements of @xmath81 and @xmath57 , finally , are described by the form factors @xmath236 where @xmath230 is defined in eq .
( [ eq : finf ] ) .
the contribution of the renormalized diagrams [ fig : o1o2]f ) , which have been omitted so far , is properly included by modifying @xmath237 as follows : @xmath238 for @xmath239 @xmath240 the loop function @xmath241 can be expanded in terms of @xmath242 .
we give the expansion of @xmath241 as well as the result for @xmath243 : @xmath244 , { \nonumber}\\ \\ h_0(0 ) & = \frac{8}{27 } -\frac{4}{9}\,\ln({\hat{s } } ) + \frac{4\,i\pi}{9 } + \frac{8}{9}\,l_\m{\nonumber}.\end{aligned}\ ] ]
the decay width differential in @xmath162 can be written as @xmath246 the last two terms in eq .
( [ rarewidth ] ) correspond to certain finite bremsstrahlung contributions specified in appendix [ appendix : brems ] .
their result can also be found in this appendix .
all other corrections have been absorbed into the effective wilson coefficients @xmath247 , @xmath248 and @xmath249 .
we follow @xcite and write the effective wilson coefficients as @xmath250 where we have provided the necessary modification to account for the ckm structure of @xmath160 . the renormalized form factors @xmath251 and @xmath252 and can be found in section [ section : renormo12u ] while the renormalized form factors @xmath253 , @xmath254 and @xmath255 are given in @xcite . the functions @xmath256 and @xmath257 encapsulate the interference between the tree - level and the one - loop matrix elements of @xmath80 and @xmath258 and the corresponding bremsstrahlung corrections , which cancel the infrared- and collinear divergences appearing in the virtual corrections .
when calculating the decay width ( [ rarewidth ] ) , we retain only terms linear in @xmath51 ( and thus in @xmath259 , @xmath260 ) in the expressions for @xmath261 , @xmath262 and @xmath263 .
accordingly , we drop terms of @xmath264 in the interference term @xmath265 , too , where by construction one has to make the replacements @xmath266 and @xmath267 in this term . the function @xmath260 has already been calculated in @xcite , where also the exact expression for @xmath268 can be found .
for the functions @xmath259 and @xmath269 and more information on the cancellation of infrared- and collinear divergences we refer to @xcite . the auxiliary quantities
@xmath270 , @xmath271 , @xmath272 , @xmath273 , @xmath274 , @xmath275 and @xmath276 are the following linear combinations of the wilson coefficients @xmath277 : @xmath278 \ , \ln \ !
\left ( \frac{m_b}{\mu } \right ) , { \nonumber}\\ a_{10 } = & \ , \frac{4\ , \pi}{\alpha_s(\mu ) } \ , c_{10}(\mu ) , \\
t_{9a } = & \ , \frac{4}{3 } \ , c_1(\mu ) + c_2(\mu)\ , , { \nonumber}\\ t_{9b } = & \ , 6 \ ,
c_3(\mu ) + 60 \ , c_5(\mu ) , { \nonumber}\\ u_9 = & - \frac{7}{2 } \ , c_3(\mu ) - \frac{2}{3 } \,c_4(\mu ) -38 \,c_5(\mu ) - \frac{32}{3 } \,c_6(\mu ) , { \nonumber}\\ w_9 = & - \frac{1}{2 } \ , c_3(\mu ) - \frac{2}{3 } \,c_4(\mu ) -8 \,c_5(\mu ) - \frac{32}{3 } \,c_6(\mu).{\nonumber}\end{aligned}\ ] ] in these definitions we also include some diagrams induced by @xmath279 insertions , viz the @xmath280 contributions , the diagrams of topology [ fig : o1o2]f ) and those bremsstrahlung diagrams where the gluon is emitted from the @xmath117- or @xmath118-quark line ( cf @xcite ) . for completeness ,
we give in table [ tablea ] numerical values for @xmath281 , @xmath282 , @xmath270 , @xmath283 , @xmath271 , @xmath272 , @xmath273 , @xmath274 , @xmath275 and @xmath276 at three different values of the renormalization scale @xmath284 .
we note that the recently calculated contributions @xcite to the anomalous dimension matrix which correspond to the three - loop mixings of the four - quark operators into @xmath81 have been included by adopting the procedure described in appendix [ appendix : wilson ] .
as can be seen in table [ tablea ] , some of the entries have a very small amount of significant digits . in our numerical analysis
presented in section [ section : phenom ] we work with a much higher accuracy .
c@c@c@c ' '' '' & @xmath285 gev & @xmath286 gev & @xmath287 gev + ' '' '' @xmath288 & @xmath289 & @xmath290 & @xmath291 + ' '' '' @xmath292 & @xmath293 & @xmath294 & @xmath295 + ' '' '' @xmath296 & @xmath297 & @xmath298 & @xmath299 + ' '' '' @xmath300 & @xmath301 & @xmath302 & @xmath303 + ' '' '' @xmath304 & @xmath305 & @xmath306 & @xmath307 + ' '' '' @xmath308 & @xmath309 & @xmath310 & @xmath311 + ' '' '' @xmath312 & @xmath313 & @xmath314 & @xmath315 + ' '' '' @xmath316 & @xmath317 & @xmath318 & @xmath319 + ' '' '' @xmath320 & @xmath321 & @xmath322 & @xmath323 + ' '' '' @xmath324 & @xmath325 & @xmath326 & @xmath327 + ' '' '' @xmath328 & @xmath329 & @xmath330 & @xmath331 +
as the main point of this paper is the _ calculation _ of the nnll corrections to the process @xmath332 , we keep the phenomenological analysis rather short . in the following
we investigate the impact of the nnll corrections on three observables : the branching ratio , the cp asymmetry and the normalized forward - backward asymmetry . as
our main point is to illustrate the differences between nll and nnll results , we do not include power corrections ( and/or effects from resonances ) , postponing this to future studies . since the decay width given in eq .
( [ rarewidth ] ) suffers from a large uncertainty due to the factor @xmath333 , we follow common practice and introduce the ratio @xmath334 in which the factor @xmath333 drops out .
note that we define @xmath335 as a charge - conjugate average as this is likely to be the first quantity measured .
the expression for the semileptonic decay width @xmath336 is as follows : @xmath337 where @xmath338 is the phase space factor , and @xmath339 incorporates the next - to - leading qcd correction to the semileptonic decay .
the function @xmath340 has been given analytically in ref .
@xcite : @xmath341 \ , .\end{aligned}\ ] ] in the following analysis we write the ckm parameters appearing in @xmath332 as ( neglecting terms of @xmath342 ) @xmath343 with @xmath344 and @xmath345 @xcite . for @xmath346 ,
appearing in the semileptonic decay width , we use @xmath347 .
numerically , we set @xmath348 , @xmath349 , @xmath350 and @xmath351 . for the other input parameters we use @xmath352 , @xmath353 gev , @xmath354 , @xmath355 gev , @xmath356 , @xmath357 gev , @xmath358 gev , and @xmath359 . in fig .
[ fig : rquark ] we show the @xmath284-dependence of @xmath360 for @xmath361 .
the solid lines correspond to the nnll results , whereas the dashed lines represent the nll results .
we see that , going from nll to nnll precision , @xmath360 is decreased throughout the entire region by about @xmath362 .
although the absolute uncertainty due to the @xmath284-dependence decreases as well , the relative error remains roughly the same . as defined in eq .
( [ eq : rquark ] ) .
the solid lines show the nnll result for @xmath363 gev , whereas the dashed lines show the corresponding result in the nll approximation . at @xmath364
the highest ( lowest ) curve correspond to @xmath365 gev ( @xmath3662.5 gev ) both in the nll and nnll case.,width=283 ] as mentioned already in the introduction , the region @xmath367 is free of resonances , as it lies below the @xmath368 threshold and above the @xmath12 and @xmath13 resonances .
the contribution of this region to the decay width ( normalized by @xmath369 ) is therefore well approximated by integrating @xmath360 over this interval . at nnll precision
, we get @xmath370 the error is obtained by varying the scale @xmath284 between @xmath371 gev and @xmath372 gev .
the corresponding result in nll precision is @xmath373 .
the renormalization scale dependence therefore increases from @xmath374 to @xmath375 .
the reason for this increace can be understood from fig .
[ fig : rquark ] : while for @xmath376 the @xmath284 dependence of @xmath377 at nnll and nll precision is similar , the @xmath284 dependence almost cancels in the nll case when integrating @xmath162 between 0.05 and 0.13 due to the crossing of the dashed lines in this interval .
this cancellation does not happen in the nnll case , leading to a slightly larger @xmath378dependence of @xmath379 at nnll .
gev , whereas the dashed lines show the corresponding result in the nll approximation . at @xmath364
the highest ( lowest ) curve correspond to @xmath365 gev ( @xmath3662.5 gev ) both in the nll and nnll case.,width=283 ] as pointed out already , in the process @xmath332 the contribution of the @xmath6-quark running in the fermion loop is , in contrast to @xmath380 , not cabibbo - suppressed . as a consequence ,
cp violation effects are much larger in @xmath332 .
the cp asymmetry @xmath381 is defined as @xmath382 in fig .
[ fig : cp ] we show @xmath381 for @xmath383 .
the solid and dashed lines correspond to the nnll and nll results , respectively .
we find several differences between the two results : the solid lines are much closer together .
also they cross each other at @xmath384 .
furthermore , the nll result clearly shows a positive cp asymmetry throughout the entire @xmath162 region considered , while the nnll lines indicate that @xmath381 can be both positive and negative , depending on the value of @xmath162 .
because of that , it does not make much sense to quantify the relative error due to the @xmath385-dependence .
the plot , however , clearly shows that the absolute uncertainty is much smaller in the nnll result . for nll results ,
see also ref .
+ we also give the averaged cp asymmetry @xmath386 in the region @xmath387 , defined as @xmath388 varying @xmath284 between 2.5 gev and 10 gev one obtains the ranges @xmath389 we now turn to the forward - backward asymmetry . as for @xmath360 , we introduce a cp - averaged version of the normalized forward - backward asymmetry , defined as @xmath390 where @xmath391 is the angle between the three - momenta of the positively charged lepton @xmath392 and the @xmath117-quark in the rest frame of the lepton pair .
the result of the integrals in the numerator of eq .
( [ eq : fbasym ] ) for the case @xmath380 can be found in @xcite .
the corresponding result for @xmath332 is , up to different ckm - structures , the same .
gev , whereas the dashed lines represent the corresponding result in the nll approximation.,width=283 ] in fig .
[ fig : fbasym ] we illustrate the @xmath284-dependence of @xmath393 in the region @xmath394 .
again , the solid and dashed lines represent the nnll and the nll results , respectively .
the reduction of the @xmath385-dependence going from nll to nnll precision is striking : one can clearly distinguish the three dashed lines , whereas the nnll lines are on top of each other throughout the region .
the position @xmath395 at which the forward - backward asymmetries vanish , is essentially free of uncertainties due to the variation of @xmath385 : we find @xmath396 . to nll precision
we get @xmath397 . as a last illustration
, we show in fig .
[ fig : matching ] the dependence of @xmath360 on the matching scales . in all the previous plots we used a matching scale of @xmath398 gev for the top contribution and a matching scale of @xmath399 gev for the charm contribution . in fig .
[ fig : matching ] the solid line corresponds to this scheme , while the dashed line is obtained by matching both contributions at a scale of @xmath399 gev .
the difference between the two schemes is between 2% and 4% . for @xmath400 gev .
the solid line corresponds to matching top and charm contributions at @xmath398 gev and @xmath399 gev , respectively .
the dashed curve is obtained by matching both contributions at a scale of @xmath399 gev.,width=283 ]
in this paper we presented the calculation of virtual and bremsstrahlung corrections of @xmath401 to the inclusive semileptonic decay @xmath402 .
genuinely new calculations were necessary to attain the virtual contributions of the operators @xmath4 and @xmath5 .
some of the diagrams ( in particular diagrams [ fig : o1o2]d ) turned out to be more involved than the corresponding diagrams for the @xmath119-quark contributions .
we used dimension - shifting and integration - by - parts techniques to calculate them . the main result of this paper , namely the @xmath6-quark contributions to the renormalized form factors @xmath403 , @xmath404 , @xmath151 , and @xmath152 ,
is given in section [ section : renormo12u ] .
we shortly discussed the numerical impact of our results on various observables in the region @xmath367 , which is known to be free of resonances .
as an example , we found the improvement on the forward - backward asymmetry @xmath405 defined in eq .
( [ eq : fbasym ] ) to be striking : the nnll result is almost free of uncertainties due to the @xmath284-dependence .
this work is partially supported by : the swiss national foundation ; rtn , bbw - contract no .
01.0357 and ec - contract hprn - ct-2002 - 00311 ( euridice ) ; nfsat - ph 095 - 02 ( crdf 12050 ) ; scopes 7ampj062165 .
we follow ref . @xcite and
derive a method that allows to express tensor integrals in @xmath406 dimensions in terms of scalar integrals of higher dimensions .
an arbitrary @xmath77 loop tensor integral with @xmath407 internal and @xmath408 external lines can be written as a linear combination of integrals of the form ( suppressing lorentz indices of @xmath409 ) @xmath410 where @xmath411 @xmath412 and @xmath413 denote the loop and external momenta , respectively .
the matrices of incidences of the diagram , @xmath13 and @xmath414 , have matrix elements @xmath415 .
the quantities @xmath416 and @xmath417 denote a set of scalar invariants formed from the external momenta @xmath413 and a set of squared masses of the internal particles , respectively .
generically , the exponents @xmath418 are equal to 1 .
however , often two or more internal lines are equipped with the same propagator .
this may be taken into account by reducing @xmath407 to @xmath419 , thus increasing some of the exponents @xmath418 . applying the integral representations
@xmath420 \\ \intertext{and } \label{eq : tensor2 } \prod\limits_{l=1}^{n_j } k_j^{\mu_{jl } } = & \ , ( -i)^{n_j}\prod\limits_{l=1}^{n_j } \frac{\partial}{\partial ( a_j)_{\mu_{jl}}}\,\exp\!\big[i(a_j\,k_j)\big ] \bigg|_{a_j=0}\end{aligned}\ ] ] allows us to easily perform the integration over the loop momenta by using the @xmath406 dimensional gaussian integration formula @xmath421 = i \left(\frac{\pi}{i\,a}\right)^{\frac{d}{2}}\ , \exp\ ! \left[-\frac{i\,p^2}{a } \right].\ ] ] we find the following parametric representation : @xmath422^\frac{d}{2}}\ \exp\!\left[\frac{i\,q\big(\{\bar s_i\},\alpha\big)}{d(\alpha ) } -i\sum\limits_{r=1}^{n } \alpha_r \left(m_r^2-i\,\epsilon \right ) \right ] \bigg|_{a_j=0}\,\,.\end{gathered}\ ] ] the quantities @xmath423 are scalar invariants involving the external momenta @xmath424 and the auxiliary momenta @xmath425 .
@xmath426 arises from the integral representations of the propagators : let @xmath427 be the @xmath77-dimensional vector that consists of all four - momentum loop vectors .
the product of all @xmath428 can then be written as @xmath429,\end{aligned}\ ] ] with @xmath412-independent quantities @xmath430 and @xmath119 .
@xmath426 denotes the determinant of the @xmath431 matrix @xmath19 . the differentiation of @xmath409 in eq .
( [ eq : parametric ] ) with respect to @xmath432 generates products of external momenta , metric tensors @xmath433 and polynomials @xmath434 and provides an additional factor @xmath435 . because of @xmath436 = r(i \partial ) \exp\!\left [ -i \sum\limits_{r=1}^{n } \alpha_r m_r^2 \right ] , \quad\text{with}\quad \partial_j = \frac{\partial}{\partial m_j^2}\ , , \ ] ] we may replace the polynomials @xmath434 with @xmath437 . the additional factor of @xmath438 can be absorbed by a redefinition of @xmath406 , i.e. by shifting @xmath406 to @xmath439 and multiplying with a factor @xmath440 .
the crucial point is that all factors generated by differentiation with respect to @xmath432 may be written as operators which do not depend on the integral representations we have introduced in eqs .
( [ eq : tensor1 ] ) , ( [ eq : tensor2 ] ) .
therefore , it is possible to write tensor integrals in momentum space in terms of scalar ones without direct appeal to the parametric representation ( [ eq : parametric ] ) : @xmath441 where the tensor operator @xmath442 ( suppressing its lorentz indices ) is given by @xmath443\ , \\ \times \prod\limits_{j=1}^{n}\ , \prod\limits_{l=1}^{n_j } \frac{\partial}{\partial ( a_j)_{\mu_{jl } } } \exp\!\big[i\ , q\big(\{\bar s_i\},\alpha\big ) \left(4\,i\pi\right)^l \bm{d^+}\big ] \bigg|_{\substack{a_j=0 \\
\alpha_j = i \partial_j}}.\end{gathered}\ ] ] the operator @xmath444 shifts the space - time dimension of the integral by two units : @xmath445 notice that throughout the derivation of the tensor operator @xmath442 the masses @xmath446 must be kept as different parameters .
they are set to their original values only in the very end . according to general rules of @xmath406 dimensional integration , integrals of the form @xmath447 vanish .
there may exist suitable linear combinations @xmath448 that lead to recurrence relations connecting the original integral to simpler ones .
the task of finding such recurrence relations , however , is in general a nontrivial one .
a criterion for irreducibility of multi - loop feynman integrals is presented in @xcite . in @xcite , the method of partial integration
is combined directly with the technique of reducing tensor integrals by means of shifting the space - time dimension .
+ the integral @xmath449^{\nu_1 } \big[r^2\big]^{\nu_2 } \big[(l+r)^2\big]^{\nu_3 } \big[(l+q)^2\big]^{\nu_4 } \big[(r+p)^2-m_b^2\big]^{\nu_5 } } \label{eq : integral}\end{gathered}\ ] ] enters the calculation of diagrams [ fig : o1o2]d ) . at the same time it is a very good example to illustrate the integration by parts method .
the operators @xmath450 , @xmath451 , ... are defined through @xmath452 the present case is especially simple because we only need to calculate one derivative . using the shorthand notation
@xmath453 we get ( for @xmath454 ) : @xmath455 i_{\{\nu_i\}}.\ ] ] scalar products of the form @xmath456 we write as @xmath457/2 $ ] and find @xmath458 i_{\{\nu_i\}}.\ ] ] at this stage we might also reduce some of the scalar products by shifting the dimension .
the corresponding procedure is presented e.g. in @xcite . in the present case , however , the pure integration by parts approach suffices .
the identity @xmath459 yields directly the desired recurrence relation for the integral @xmath460 : @xmath461 subsequent application of this relation allows to express any integral @xmath462 with indices @xmath463 as a sum over integrals @xmath462 with at least @xmath464 or @xmath465 .
the general procedure is the following : * one expresses suitable scalar products in the numerator of a given feynman integrand in terms of inverse propagators @xmath466 and cancels them down .
it is important to notice that it is not always the best strategy to try to cancel down as many scalar products as possible .
the resulting set of integrals to calculate highly depends on which scalar products one cancels down .
the best way is to try a couple of different cancellation schemes and compare the resulting integrals .
* one writes the integral as a sum over tensor integrals of the form ( [ eq : gd ] ) with products of @xmath467 . for each of those integrals
the tensor operator @xmath442 is determined in order to reduce the problem to scalar integrals with shifted space - time dimension . *
one applies appropriate recurrence relations to reduce the number of propagators in the integrals , hoping to be able to solve the remaining integrals .
the contribution of the sum of diagrams [ fig : o1o2]d ) is given by a combination of integrals of the form @xmath468^{\nu_1 } \big[r^2\big]^{\nu_2 } \big[(l+r)^2\big]^{\nu_3 } \big[(l+q)^2\big]^{\nu_4 } \big[(r+p)^2-m_b^2\big]^{\nu_5 } } \,.\end{aligned}\ ] ] in this section we show how to solve these integrals with the methods presented in appendices [ section : tensor ] and [ section : intbyparts ] .
the function @xmath469 , which is independent of @xmath470 and @xmath471 , is not needed in order to find the tensor operators @xmath442 .
nevertheless , we give it as an illustration : @xmath472 the function @xmath473 , however , must be calculated for each type of tensor integral . as an example we give @xmath474 for @xmath475 , @xmath476 : @xmath477 the corresponding tensor operator
@xmath442 reads : @xmath478.\ ] ] the action of an operator @xmath479 on the integral @xmath480 is @xmath481 the next step is to repeatedly apply the recurrence relation ( [ eq : recrel ] ) on the integrals @xmath460 until @xmath482 or @xmath483 becomes zero .
the problem is then reduced to the calculation of the two types of integrals @xmath484 in the present calculation @xmath406 may take the values @xmath485 it is important to note here that the denominator in eq .
( [ eq : recrel ] ) can become proportional to @xmath486 for certain values of @xmath487 and @xmath488 .
thus , some of the integrals in ( [ eq : types ] ) need to be calculated up to @xmath489 . the first type of integrals ( @xmath490 ) can easily be solved individually by using a single mellin - barnes approach .
this method naturally results in an expansion in @xmath162 .
furthermore , the occasionally needed @xmath489 terms are easily obtained since the expansion in @xmath486 is done only in the very end .
we now turn to the much more complicated calculation of the second set of integrals . instead of calculating every single occurring integral individually , we derive a general formula for @xmath491 where we are left with a three - dimensional feynman parameter integral : @xmath492 we now replace all occurrences of @xmath491 according to eq .
( [ eq : generic2 ] ) and are left with a three - dimensional integral over a rather lengthy integrand .
this integrand can be split up into three different parts : * a part with no additional divergences arising from the integrations . * a part with problematic @xmath493-integration . *
a part with problematic @xmath6-integration . in the first part
, the regulator @xmath494 is not needed at all and may be set equal to zero at the very beginning .
the occurring integrals can then either be performed directly or with the use of a single mellin - barnes representation .
the second part boils down to two different integrals , which can both be computed with subtraction methods .
the last part is clearly the most difficult one .
it can be reduced to three integrals which we calculate using a double mellin - barnes representation . since this double mellin - barnes is very different from the one presented in subsection 3.1.4 in @xcite , we give , as an example , the needed procedure to calculate one of the three integrals .
specifically , we have to deal with the integrals @xmath495 where @xmath184 can take the values @xmath496 and @xmath497 .
we focus on the case where @xmath498 .
we introduce a first mellin - barnes integral in the complex @xmath62-plane with the identifications ( for notation see e.g. @xcite ) : @xmath499 and get @xmath500 the path @xmath501 lies in the left half - plane and can be chosen arbitrarily close to the imaginary @xmath62-axis .
we introduce a second mellin - barnes representation in the complex @xmath502-plane for the last factor in the denominator of eq .
( [ eq : mb1 ] ) . for this , we rewrite @xmath503 as @xmath504 and make the following identifications : @xmath505 yielding @xmath506 the path @xmath507 lies to the left of the imaginary @xmath502-axis and can again be chosen arbitrarily close to that axis .
the parameter integrals can now be performed and give products of euler beta - functions .
we work out the remaining integrals over @xmath62 and @xmath502 applying the residue theorem . for this , we close the @xmath62-integral in the right half - plane and focus on the enclosed poles .
there are two different sequences of poles , namely poles that depend on @xmath502 ( coupled poles ) and poles that do not ( uncoupled poles ) .
the latter poles lie at the following positions : * @xmath508 * @xmath509 note here that @xmath510 exists only for negative values of @xmath486 .
the pole located at @xmath511 therefore lies in the right half - plane and needs to be taken into account .
since we are interested in an expansion in @xmath162 , we can truncate the two pole sequences at a suitable @xmath512 . after calculating the necessary residues ,
we close the @xmath502-integral in the right half - plane as well and are arrive at pole sequences situated at the following positions : * @xmath513 * @xmath514 * @xmath515 for @xmath516 + @xmath517 for @xmath518 for @xmath519 , some of the poles above lie in the left @xmath502-half - plane and must be omitted . unlike the procedure given in subsection 3.1.4 of @xcite
, we need to sum up the residues of all poles in the enclosed area .
calculating the contributions of the coupled poles in @xmath62 , which lie at @xmath520 , yields an expression that is proportional to @xmath521 .
two problems now arise if one closes the integration path of the @xmath502-integral in the right half - plane : due to the @xmath522 in the exponent of @xmath162 , one gets an expansion in inverse powers of @xmath162 , forcing one to calculate the residues of all enclosed poles .
the second problem is even worse : for any given value of @xmath523 , there always exists an infinite pole series in @xmath502 which contributes to the desired result .
thus , one also has to consider the infinite pole series in @xmath62 . in order to avoid these problems
, we close the integration path in the left half - plane of @xmath502 .
the poles are then located at * @xmath524 * @xmath525 * @xmath526 after calculating the necessary residues we obtain the result for @xmath510 .
the results for @xmath527 and @xmath528 are calculated in an analogous way .
the wilson coefficients satisfy the renormalization group equation @xmath529 where @xmath97 is the anomalous dimension matrix .
this matrix can be written as a taylor series in @xmath530 : @xmath531 the general solution of eq .
( [ eq : rge1 ] ) can be expressed with the evolution matrix @xmath532 : @xmath533 the aim in this section is to find a handy expression for @xmath534 .
+ the matrix @xmath535 can be diagonalized .
we introduce new quantities in the following way : @xmath536 the matrix @xmath537 is chosen such that @xmath538 is diagonal .
one can check that the new quantities satisfy equations similar to ( [ eq : rge1 ] ) and ( [ eq : evolution ] ) : @xmath539 we will now construct a solution to eq .
( [ eq : rge2 ] ) .
once this solution is found , we can easily gain the solution of the initial problem for the non - diagonal @xmath535 .
the evolution matrix @xmath540 satisfies the same equation as @xmath541 itself : @xmath542 we make the following ansatz for @xmath543 : @xmath544 where @xmath545 solves eq .
( [ eq : umatrix ] ) to leading logarithmic approximation and is given by @xmath546 the vector @xmath547 collects the diagonal elements of @xmath538 .
the matrix @xmath548 must be chosen such that the boundary condition given in eq .
( [ eq : rge2 ] ) is met .
the quantities @xmath549 appear in the rge for @xmath530 : @xmath550 inserting the ansatz ( [ eq : ansatz ] ) into eq .
( [ eq : umatrix ] ) and using the explicit expression for @xmath545 , the lhs and the rhs of this equation can be written as @xmath551 the unknown matrices @xmath552 can now be constructed order by order in @xmath530 through the relations @xmath553 .
we give the explicit solutions to @xmath554 and @xmath555 since we need them to find the wilson coefficients @xmath556 to nnll precision : @xmath557 the result for @xmath554 agrees with the one given in section iii of @xcite . after we did the calculation for @xmath555 , we found out that the result already exists in the literature @xcite .
the two results agree as well .
+ the matrix @xmath548 is given through @xmath558 with these informations at hand , we can present the evolution matrix for the initial problem given in eqs .
( [ eq : rge1 ] ) and ( [ eq : evolution ] ) : @xmath559
in order to fix the counterterms @xmath560 ( @xmath168 ) in eq .
( [ fi4quark ] ) , we need the one - loop matrix elements @xmath561 of the four - quark operators @xmath4 , @xmath5 , @xmath562 , @xmath563 and @xmath564 . due to the @xmath565 factor in eq .
( [ fi4quark ] ) they are needed up to @xmath566 .
the explicit results read @xmath567 + \\ & \frac{\epsilon}{81 } \big[\ , 52 + 24\,i\pi - 21\,\pi^2 - ( 24 + 36\,i\pi ) l_s + 18\,l_s^2\ , \big]\ ! \bigg\}\ , \bra \widetilde{o}_9\ket_{\text{tree}}\ , , \\ \\
\bra d\,\ell^+\ell^- |o_1^u|b\ket_{\text{1-loop } } = & \ , \frac{4}{3}\ , \bra d\,\ell^+\ell^- |o_2^u|b\ket_{\text{1-loop } } \ , , \\ \\
\bra d\,\ell^+\ell^- |o_4|b\ket_{\text{1-loop } } = & -\left(\frac{\mu}{m_b}\right)^{2\epsilon } \bigg\{\ ! \left[\frac{4}{9
} + \frac{\epsilon}{945 } \left(70\ , { \hat{s}}+ 7\ , { \hat{s}}^2 + { \hat{s}}^3\right)\right ] \ , \bra \widetilde{o}_7\ket_{\text{tree } } \\ & + \bigg [ \frac{16}{27\,\epsilon } + \frac{2}{8505 } \left ( - 420 + 1260\ , i\pi - 1260\ , l_s + 252\ , { \hat{s}}+ 27\ , { \hat{s}}^2 + 4\ , { \hat{s}}^3 \right ) \\ & + \frac{4\,\epsilon } { 8505 } \left(420\ , i\pi + 910 - 630\,l_s\,i\pi - 420\ , l_s - 315\,\pi^2 \right . \\ & + \left . \left .
315\,l_s^2 - 126\,{\hat{s}}+ { \hat{s}}^3 \right)\bigg]\bra \widetilde{o}_9\ket_{\text{tree}}\right\ } , \\\end{aligned}\ ] ] @xmath568
in section [ section : decay_width ] those bremsstrahlung contributions were taken into account which generate infrared and collinear singularities .
combined with virtual contributions which also suffer from such singularities , a finite result was obtained . in this appendix
we discuss the remaining finite bremsstrahlung corrections which are encoded in the last two terms of eq .
( [ rarewidth ] ) .
being finite , these terms can be directly calculated in @xmath174 dimensions .
the sum of the bremsstrahlung contributions from @xmath569 and @xmath570 interference terms and the @xmath571 term can be written as @xmath572 + c_{88}\,\tau_{88 } \big),\end{gathered}\ ] ] where @xmath573 for the quantities @xmath574 , @xmath575 and @xmath576 we refer to @xcite .
the remaining bremsstrahlung contributions all involve the diagrams with an @xmath48 or @xmath49 insertion where the gluon is emitted from the @xmath6- or @xmath119-quark loop , respectively .
the corresponding bremsstrahlung matrix elements depend on the functions @xmath577 , @xmath578 . in @xmath174 dimensions we find @xmath579 where @xmath580 + the analytical expressions for @xmath581 and @xmath582 can be written in terms of functions @xmath583 : @xmath584 , \\
{ \nonumber}\\ \bar\delta i_{27}^{(c ) } & = 2 \left [ g_0\!\left(\frac{{\hat{s}}}{z}\right ) - g_0\!\left(\frac{w}{z}\right ) \right],\end{aligned}\ ] ] where @xmath585 .
@xmath586 ( @xmath587 ) is defined through the integral @xmath588 , \quad g_1(t ) = \frac{1}{2 } g_0(t).\ ] ] explicitly , the functions @xmath589 and @xmath590 read @xmath591 the quantities @xmath592 we obtain from @xmath593 in the limit @xmath594 : @xmath595 \ , , \\
\bar\delta i_{27}^{(u ) } & = -2 \big [ \ln(w)- \ln({\hat{s } } ) \big].\end{aligned}\ ] ] following @xcite , we write @xmath596 \big\}.\end{gathered}\ ] ] expressed in terms of the quantities @xmath597 and @xmath598 , defined by @xmath599 the quantities @xmath600 introduced in eq .
( [ eq : tau ] ) read @xmath601 \left| \bar\delta i_{23}^{\text{eff}}\right|^2 + \\ \big [ 2\,{\hat{s}}^2\,(2+\,w ) + { \hat{s}}\,w\,(1 + 2\,w ) \big ] \left| \bar\delta i_{27}^{\text{eff}}\right|^2 + 4\,{\hat{s}}\big[w\,(1-w ) - { \hat{s}}\,(2+w ) \big ] \cdot \text{re}\left [ \bar\delta i_{23}^{\text{eff}}\bar\delta i_{27}^{{\text{eff } } * } \right ] \!\!\bigg\},\end{gathered}\ ] ] @xmath603 \bar\delta i_{23}^{\text{eff}}- \big [ 2\,{\hat{s}}\,(w-{\hat{s}})^2(1-w ) \big ] \bar\delta i_{27}^{\text{eff}}\\ + { \hat{s}}\,w \big [ ( 1 + 2\,{\hat{s}}-2\,w)\bar\delta i_{23}^{\text{eff}}- 2\ , ( 1+{\hat{s}}-w ) \bar\delta i_{27}^{\text{eff}}\big ] \cdot \ln\!\left [ \frac{{\hat{s}}}{(1+{\hat{s}}-w)(w^2+{\hat{s}}\,(1-w ) ) } \right ] \!\!\bigg\},\end{gathered}\ ] ] the coefficients @xmath605 include the dependence on the wilson coefficients and the color factors .
@xmath606 , & c_{18 } = & \ , c_{\tau_2 } \cdot c_1^{(0 ) } { \widetilde{c}}_8^{(0,{\text{eff } } ) * } , & c_{28 } = & \ , c_f \cdot c_2^{(0 ) } { \widetilde{c}}_8^{(0,{\text{eff } } ) * } , \\
c_{22 } = & \ , c_f \cdot \left|c_2^{(0)}\right|^2 , & c_{19 } = & \ , c_{\tau_2 } \cdot c_1^{(0 ) } { \widetilde{c}}_9^{(0,{\text{eff } } ) * } , & c_{29 } = & \ , c_f \cdot c_2^{(0 ) } { \widetilde{c}}_9^{(0,{\text{eff}})*}. { \nonumber}\end{aligned}\ ] ] the wilson coefficients @xmath607 are given in eq .
( [ effcoeff ] ) and numerical values for the coefficients @xmath91 can be found in table [ tablea ] .
the color factors @xmath608 , @xmath609 and @xmath610 are presented in section [ section : eff ] .
j. kaneko _ et al .
_ [ belle collaboration ] , phys .
* 90 * ( 2003 ) 021801 , .
k. abe _ et al .
_ [ belle collaboration ] , .
b. aubert _ et al . _ [ babar collaboration ] , .
z. ligeti and m. b. wise , phys . rev . *
d 53 * ( 1996 ) 4937 , .
a. f. falk , m. luke and m. j. savage , phys .
d 49 * ( 1994 ) 3367 , .
a. ali , g. hiller , l. t. handoko and t. morozumi , phys . rev . *
d 55 * ( 1997 ) 4105 , .
chen , g. rupak and m. j. savage , phys .
* b 410 * ( 1997 ) 285 , .
g. buchalla , g. isidori and s. j. rey , nucl .
* b 511 * ( 1998 ) 594 , . g. buchalla and g. isidori , nucl . phys . * b 525 * ( 1998 ) 333 , . f. kruger and l. m. sehgal , phys . rev . *
d 55 * ( 1997 ) 2799 , .
a. ali , p. ball , l.t .
handoko , g. hiller , phys . rev .
* d 61 * ( 2000 ) 074024 , .
e. lunghi and i. scimemi , nucl
* b 574 * ( 2000 ) 43 , .
e. lunghi , a. masiero , i. scimemi and l. silvestrini , nucl . phys . * b 568 * ( 2000 ) 120 , .
k. chetyrkin , m. misiak and m. mnz , phys . lett .
* b 400 * ( 1997 ) 206 , .
m. ciuchini , e. franco , g. martinelli , l. reina and l. silvestrini , phys . lett .
* b 316 * ( 1993 ) 127 , ; nucl . phys . * b 415 * ( 1994 ) 403 , ; g. cella , g. curci , g. ricciardi and a. vicere , phys . lett . *
b 325 * ( 1994 ) 227 , .
k. adel and y .- p .
yao , phys . rev . *
d 49 * ( 1994 ) 4945 , . c. greub and t. hurth , phys
d 56 * ( 1997 ) 2934 , .
buras , a. kwiatkowski and n. pott , nucl .
* b 517 * ( 1998 ) 353 , ; phys .
* b 414 * ( 1997 ) 157 , , e : ibid .
* 434 * ( 1998 ) 459 .
a. ali and c. greub , z. phys .
c * 49 * ( 1991 ) 431 ; phys .
* b 259 * ( 1991 ) 182 .
n. pott , phys . rev .
* d 54 * ( 1996 ) 938 , . c. greub , t. hurth and d. wyler , phys .
* b 380 * ( 1996 ) 385 , ; phys . rev . * d 54 * ( 1996 ) 3350 , .
a. ali , h. asatrian and c. greub , phys .
* b 429 * ( 1998 ) 87 , .
t. hurth and t. mannel , aip conf .
* 602 * ( 2001 ) 212 , .
t. hurth and t. mannel , phys . lett . * b 511 * ( 2001 ) 196 , . c. bobeth , m. misiak and j. urban , nucl
* b 574 * ( 2000 ) 291 , .
gambino , m. gorbahn and u. haisch , nucl .
* b 673 * ( 2003 ) 238 , .
h. h. asatryan , h. m. asatrian , c. greub and m. walker , phys . rev .
* d 65 * ( 2000 ) 074004 , .
h. h. asatryan , h. m. asatrian , c. greub and m. walker , phys . rev . *
d 66 * ( 2002 ) 034009 , .
a. ghinculov , t. hurth , g. isidori and y. p.
a. ghinculov , t. hurth , g. isidori and y. p. yao , nucl . phys .
* b 648 * ( 2003 ) 254 , .
h. m. asatrian , k. bieri , c. greub and a. hovhannisyan , phys .
d 66 * ( 2002 ) 094013 , .
a. ghinculov , t. hurth , g. isidori and y. p. yao , nucl .
phys . proc .
suppl . * 116 * ( 2003 ) 284 , .
h. m. asatrian , h. h. asatryan , a. hovhannisyan and v. poghosyan , . t. hurth , rev .
phys . * 75 * ( 2003 ) 1159 , .
o. v. tarasov , phys .
d 54 * ( 1996 ) 6479 , .
a. baikov , phys . lett .
* b 474 * ( 2000 ) 385 , .
m. misiak , nucl . phys .
* b 393 * ( 1993 ) 23 , nucl .
* b 439 * ( 1995 ) 461 ( e ) .
g. buchalla , a. j. buras and m. e. lautenbacher , rev .
* 68 * ( 1996 ) 1125 , .
a. j. buras and m. munz , phys . rev .
* d 52 * ( 1995 ) 186 , .
b. grinstein , m. j. savage and m. b. wise , nucl .
* b 319 * ( 1989 ) 271 .
v. a. smirnov , _ renormalization and asymptotic expansions _
( birkhuser , basel , 1991 ) ; _ applied asymptotic expansions in momenta and masses _ ( springer - verlag , heidelberg , 2001 ) ; mod .
lett . a * 10 * ( 1995 ) 1485 , . o. v. tarasov , nucl . phys . *
b 562 * ( 1997 ) 455 , .
h. h. asatryan , h. m. asatrian , c. greub and m. walker , phys . lett .
* b 507 * ( 2001 ) 162 , .
y. nir , phys .
* b 221 * ( 1989 ) 184 .
a. j. buras , m. e. lautenbacher and g. ostermaier , phys . rev . *
d 50 * ( 1994 ) 3433 , .
a. ali and g. hiller , eur .
j. c * 8 * ( 1999 ) 619 , .
m. beneke , t. feldmann and d. seidel , nucl . phys .
* b 612 * ( 2001 ) 25 , . | we present the calculation of the virtual- and bremsstrahlung corrections of @xmath0 to the matrix elements @xmath1 .
this is the missing piece in the full next - to - next - to - leading logarithmic ( nnll ) results for various observables associated with the process @xmath2 , like the branching ratio , the cp - rate asymmetry and the forward - backward asymmetry .
this paper is an extension of analogous calculations done by some of us for the process @xmath3 .
as the contributions of the diagrams induced by the four - quark operators @xmath4 and @xmath5 with a @xmath6-quark running in the quark loop are strongly ckm suppressed , they were omitted in the analysis of @xmath3 .
this is no longer possible for @xmath7 , as the corresponding contributions are not suppressed .
the main new work therefore consists of calculating the @xmath8 corrections to @xmath9 . in this paper we restrict ourselves to the range @xmath10 ( @xmath11 is the invariant mass of the lepton pair ) , which lies above the @xmath12- and @xmath13-resonances and below the @xmath14-resonance .
we present the analytic results for the mentioned observables related to the process @xmath15 as expansions in the small parameters @xmath16 , @xmath17 and @xmath18 . in the phenomenological analysis at the end of the paper
we discuss the impact of the nnll corrections on the observables mentioned above . |
in this section we describe the main results of the paper .
( in case any of the terminology appears to be new to the reader , the corresponding definitions are given in the next section . )
if a contact structure on a @xmath0-manifold is cooriented , then every legendrian knot ( i.e. a knot that is everywhere tangent to the contact distribution ) has a natural framing ( a continuous normal vector field ) . hence when studying legendrian knots in such contact manifolds the main question is to distinguish those of them that realize isotopic framed knots .
similarly if the contact structure is parallelized , then every transverse knot ( i.e. a knot that is everywhere transverse to the contact distribution ) also has a natural framing , and when studying transverse knots in such contact manifolds again the main question is to distinguish those of them that realize isotopic framed knots .
vassiliev invariants proved to be an extremely useful tool in the study of framed knots , and the conjecture is that they are sufficient to distinguish all the isotopy classes of framed knots .
vassiliev invariants can also be easily defined in the categories of legendrian and of transverse knots . in this paper
we study the relationship between the groups of vassiliev invariants of these three categories of knots , and explore when these invariants can be used to distinguish legendrian knots that realize isotopic framed knots .
consider a contact manifold @xmath4 with a cooriented contact structure .
fix an abelian group @xmath5 , a connected component @xmath6 of the space of framed immersions of @xmath7 into @xmath4 , and a connected component @xmath8 of the space of legendrian immersions of @xmath7 into @xmath4 .
we study the relation between the groups of @xmath5-valued vassiliev invariants of framed knots from @xmath6 and of @xmath5-valued vassiliev invariants of legendrian knots from @xmath9 .
the main results obtained in this paper are described below .
[ first ] the groups of @xmath5-valued vassiliev invariants of legendrian knots from @xmath9 and of framed knots from @xmath6 are canonically isomorphic , provided that the euler class of the contact bundle vanishes on every @xmath10 realizable by a mapping @xmath11 .
( see theorem [ isomorphismobtained ] and proposition [ interpretationconditionii ] . )
using theorem [ first ] we show that : the groups of @xmath5-valued vassiliev invariants of legendrian knots from @xmath9 and of framed knots from @xmath6 are canonically isomorphic , provided that one of the following conditions holds : 1 : : the contact structure is tight ; 2 : : the euler class of the contact bundle is in the torsion of @xmath12 ( in particular if the euler class is zero ) .
3 : : the contact manifold is closed and admits a metric of negative sectional curvature .
( see sections [ homologysphere ] and [ negativecurvature ] and theorem [ tight ] . ) as a corollary , we get that for any surface @xmath2 the group of finite order arnold s @xmath1-type invariants of wave fronts on @xmath2 is isomorphic to the group of vassiliev invariants of framed knots in the spherical cotangent bundle @xmath3 of @xmath2 . previously the isomorphism of the groups of vassiliev invariants of legendrian and of framed knots was known only in the case where @xmath13 and @xmath4 is the standard contact @xmath14 ( result of d. fuchs and s. tabachnikov @xcite ) or the standard contact solid - torus ( result of j. hill @xcite ) .
the proofs of these isomorphisms were based on the fact that for the @xmath15-valued vassiliev invariants of framed knots in these manifolds there exists a universal vassiliev invariant also known as the kontsevich integral .
( currently the existence of the kontsevich integral is known only for a total space of an @xmath16-bundle over a compact oriented surface with boundary , see the paper @xcite of andersen , mattes , and reshetikhin . )
thus the approach used in @xcite and @xcite to show the isomorphism of the groups of vassiliev invariants is not applicable for almost all contact @xmath0-manifolds and abelian groups @xmath5 and our results appear to be a strong generalization of the results of fuchs , tabachnikov and hill .
we also construct the first examples where vassiliev invariants can be used to distinguish legendrian knots that realize isotopic framed knots and are homotopic as legendrian immersions .
these are also the first examples where the groups of vassiliev invariants of legendrian and of framed knots from the corresponding components of the spaces of legendrian and of framed immersions are not canonically isomorphic .
the manifold @xmath17 admits infinitely many cooriented contact structures for which there exist legendrian knots that can be distinguished by @xmath18-valued vassiliev invariants even though they realize isotopic framed knots and are homotopic as legendrian immersions .
( see theorem [ example1 ] and theorem [ example2 ] in which the similar result is proved for any orientable total space of an @xmath7-bundle over a nonorientable surface of a sufficiently high genus . ) for transverse knots we obtain the following result ( see theorem [ isomorphismobtainedtransverse ] ) : let @xmath19 be a contact manifold with a parallelized contact structure , then the groups of @xmath5-valued vassiliev invariants of transverse and of framed knots ( from the corresponding components of the spaces of transverse and of framed immersions ) are canonically isomorphic .
in this paper @xmath5 is an abelian group ( not necessarily torsion free ) , and @xmath4 is a connected oriented @xmath0-dimensional riemannian manifold ( not necessarily compact ) .
a _ contact structure _ on a @xmath0-dimensional manifold @xmath4 is a smooth field @xmath20 of tangent @xmath21-dimensional planes , locally defined as a kernel of a differential @xmath22-form @xmath23 with non - vanishing @xmath24 .
a manifold with a contact structure possesses the canonical orientation determined by the volume form @xmath25 .
the standard contact structure in @xmath14 is the kernel of the @xmath22-form @xmath26 .
a _ contact element _ on a manifold is a hyperplane in the tangent space to the manifold at a point . for a surface @xmath2
we denote by @xmath3 the space of all cooriented ( transversally oriented ) contact elements of @xmath2 .
this space is the spherical cotangent bundle of @xmath2 .
its natural contact structure is the distribution of tangent hyperplanes given by the condition that the velocity vector of the incidence point of a contact element belongs to the element .
a contact structure is _ cooriented _ if the @xmath21-dimensional planes defining the contact structure are continuously cooriented ( transversally oriented ) .
a contact structure is _ oriented _ if the @xmath21-dimensional planes defining the contact structure are continuously oriented . since every contact manifold has a natural orientation we see that every cooriented contact structure is naturally oriented and every oriented contact structure is naturally cooriented .
a contact structure is _ parallelizable _ ( _ parallelized _ ) if the @xmath21-dimensional vector bundle @xmath27 over @xmath4 is trivializable ( trivialized ) .
since every contact manifold has a canonical orientation , one can see that every parallelized contact structure is naturally cooriented .
a contact structure @xmath28 on a manifold @xmath4 is said to be _ overtwisted _ if there exists a @xmath21-disk @xmath29 embedded into @xmath4 such that the boundary @xmath30 is tangent to @xmath28 while the disk @xmath29 is transverse to @xmath28 along @xmath30 .
not overtwisted contact structures are called _
tight_. a _ curve _ in @xmath4 is an immersion of @xmath7 into @xmath4 .
( all curves have the natural orientation induced by the orientation of @xmath7 . ) a _ framed curve _ in @xmath4 is a curve in @xmath4 equipped with a continuous unit normal vector field .
a _ legendrian curve _ in a contact manifold @xmath19 is a curve in @xmath4 that is everywhere tangent to @xmath28 .
if the contact structure on @xmath4 is cooriented , then every legendrian curve has a natural framing given by the unit normals to the planes of the contact structure that point in the direction specified by the coorientation . to a legendrian curve @xmath31 in a contact manifold with a parallelized contact structure one can associate an integer that is the number of revolutions of the direction of the velocity vector of @xmath31 ( with respect to the chosen frames in @xmath28 ) under traversing @xmath31 according to the orientation .
this integer is called the _ maslov number _ of @xmath31 .
the set of maslov numbers enumerates the set of the connected components of the space of legendrian curves in @xmath14 ( cf .
[ h - principlelegendrian ] ) .
a _ transverse _ curve in a contact manifold @xmath19 is a curve in @xmath4 that is everywhere transverse to @xmath28 .
if the contact structure on @xmath4 is parallelized , then a transverse curve has a natural framing given by the unit normals corresponding to the projections of the first of the two coordinate vectors of the contact planes on the @xmath21-planes orthogonal to the velocity vectors of the curve . a transverse curve in a contact manifold with a cooriented contact structures is said to be _ positive _ if at every point the velocity vector of the curve points into the coorienting half - plane , and it is said to be _ negative _ otherwise .
there are two connected components of the space of transverse curves in @xmath14 , they consist of positive and negative transverse curves respectively . in general if @xmath19 is a contact manifold with a cooriented contact structure , then every connected component of the space of unframed curves contains two connected components of the space of transverse curves .
they consist of positive and negative transverse curves respectively . a _ knot ( framed knot ) _ in @xmath4 is an embedding ( framed embedding ) of @xmath7 into @xmath4 . in a similar way we define legendrian and transverse knots in @xmath4 .
a _ singular ( framed ) _ knot with @xmath32 double points is a curve ( framed curve ) in @xmath4 whose only singularities are @xmath32 transverse double points .
an _ isotopy _ of a singular ( framed ) knot with @xmath32 double points is a path in the space of singular ( framed ) knots with @xmath32 double points under which the preimages of the double points on @xmath7 change continuously .
an @xmath5-valued framed ( resp .
legendrian , resp .
transverse ) knot invariant is an @xmath5-valued function on the set of the isotopy classes of framed ( resp .
legendrian , resp .
transverse ) knots . a transverse double point @xmath33 of a singular knot can be resolved in two essentially different ways .
we say that a resolution of a double point is positive ( resp .
negative ) if the tangent vector to the first strand , the tangent vector to the second strand , and the vector from the second strand to the first form the positive @xmath0-frame .
( this does not depend on the order of the strands ) . if the singular knot is legendrian ( resp .
transverse ) , then these resolution can be made in the category of legendrian ( resp .
transverse ) knots . a singular framed ( resp .
legendrian , resp .
transverse ) knot @xmath34 with @xmath35 transverse double points admits @xmath36 possible resolutions of the double points .
the sign of the resolution is put to be @xmath37 if the number of negatively resolved double points is even , and it is put to be @xmath38 otherwise .
let @xmath39 be an @xmath5-valued invariant of framed ( resp .
legendrian , resp .
transverse ) knots .
the invariant @xmath39 is said to be of _ finite order _ ( or _ vassiliev invariant _ ) if there exists a nonnegative integer @xmath32 such that for any singular knot @xmath40 with @xmath35 transverse double points the sum ( with appropriate signs ) of the values of @xmath39 on the nonsingular knots obtained by the @xmath36 resolutions of the double points is zero .
an invariant is said to be of order not greater than @xmath32 ( of order @xmath41 ) if @xmath32 can be chosen as the integer in the definition above .
the group of @xmath42-valued finite order invariants has an increasing filtration by the subgroups of the invariants of order @xmath41 .
[ h - principlelegendrian ] _
@xmath43-principle for legendrian curves .
_ for @xmath19 a contact manifold with a cooriented contact structure , we put @xmath44 to be the total space of the fiberwise spherization of the contact bundle , and we put @xmath45 to be the corresponding locally trivial @xmath7-fibration .
the @xmath43-principle proved for the legendrian curves by m. gromov ( @xcite , pp.338 - 339 ) says that the space of legendrian curves in @xmath19 is weak homotopy equivalent to the space of free loops @xmath46 in @xmath44 .
the equivalence is given by mapping a point of a legendrian curve to the point of @xmath44 corresponding to the direction of the velocity vector of the curve at this point .
in particular the @xmath43-principle implies that the set of the connected components of the space of legendrian curves can be naturally identified with the set of the conjugacy classes of elements of @xmath47 .
[ description]_description of legendrian and of transverse knots in @xmath14 . _
the contact darboux theorem says that every contact @xmath0-manifold @xmath19 is locally contactomorphic to @xmath14 with the standard contact structure that is the kernel of the @xmath22-form @xmath26 .
a chart in which @xmath19 is contactomorphic to the standard contact @xmath14 is called _ a darboux chart .
_ transverse and legendrian knots in the standard contact @xmath14 are conveniently presented by the projections into the plane @xmath48 .
identify a point @xmath49 with the point @xmath50 furnished with the fixed direction of an unoriented straight line through @xmath48 with the slope @xmath51 .
then the curve in @xmath14 is a one parameter family of points with ( non - vertical ) directions in @xmath52 .
a curve in @xmath14 is transverse if and only if the corresponding curve in @xmath52 is never tangent to the chosen directions along itself .
while a generic regular curve has a regular projection into the @xmath48-plane , the projection of a generic legendrian curve into the @xmath48-plane has isolated critical points ( since all the planes of the contact structure are parallel to the @xmath51-axis ) .
hence the projection of a generic legendrian curve may have cusps .
a curve in @xmath14 is legendrian if and only if the corresponding planar curve with cusps is everywhere tangent to the field of directions . in particular this field
is determined by the curve with cusps .
let @xmath19 be a contact manifold with a cooriented contact structure .
let @xmath9 be a connected component of the space of legendrian curves in @xmath4 , and let @xmath6 be the connected component of the space of framed curves that contains @xmath9 .
( such a component exists because a legendrian curve in a manifold with a cooriented contact structure is naturally framed , and a path in the space of legendrian curves corresponds to a path in the space of framed curves . )
let @xmath53 ( resp .
@xmath54 ) be the group of @xmath5-valued order @xmath41 invariants of legendrian ( resp .
framed ) knots from @xmath9 ( resp . from @xmath6 ) .
clearly every invariant @xmath55 restricted to the category of legendrian knots in @xmath9 is an element @xmath56 .
this gives a homomorphism @xmath57 .
[ isomorphism ] let @xmath58 be a contact manifold with a cooriented contact structure .
let @xmath9 be a connected component of the space of legendrian curves in @xmath4 , and let @xmath6 be the connected component of the space of framed curves that contains @xmath9 . then the following two statements * a * and * b * are equivalent . a : : @xmath59 for any @xmath60 and any knots @xmath61 representing isotopic framed knots .
b : : @xmath62 is a canonical isomorphism .
if the mapping from the isotopy classes of legendrian knots in @xmath9 to the isotopy classes of framed knots in @xmath6 is surjective , then the proof of theorem [ isomorphism ] is obvious .
however in general this mapping is not surjective and the proof of theorem [ isomorphism ] is given in section [ proofisomorphism ] .
the famous bennequin inequality shows that this mapping is not surjective even in the case where @xmath4 is the standard contact @xmath14 .
theorem [ isomorphism ] implies that to obtain the isomorphism between the groups @xmath54 and @xmath53 it suffices to show that statement @xmath63 of theorem [ isomorphism ] is true for the connected components @xmath9 and @xmath6 of the spaces of legendrian and of framed curves .
[ conditions]_condition @xmath64 . _ in @xcite d. fuchs and s. tabachnikov showed that statement @xmath63 holds for all the connected components of the space of legendrian curves when the ambient manifold is the standard contact @xmath14 and the group @xmath5 is @xmath15 .
( one can verify that the proof of this theorem of fuchs and tabachnikov goes through for @xmath5 being any abelian group . )
they later observed @xcite that since their proof of this fact is mostly local , the similar fact should be true for a big class of contact manifolds .
however in fact the proof of their theorem is not completely local and is also based on the existence of well - defined bennequin invariant and maslov number for a legendrian knot in @xmath14 .
in general the bennequin invariant is not well - defined unless the knot is zero - homologous and the maslov number is not well - defined unless either the knot is zero - homologous or the contact structure is parallelizable .
thus the generalization of this theorem to the case of manifolds other than @xmath14 meets certain difficulties .
( and in fact the corresponding result does not hold for a big class of contact manifolds , see section [ nonisomorphicgroups ] . ) by analyzing the proof of the theorem of fuchs and tabachnikov ( see section [ reasonstoappear ] ) we get that it can be generalized to the case of an arbitrary contact @xmath0-manifold with a cooriented contact structure , provided that the connected component @xmath6 ( containing @xmath9 ) satisfies the following * condition @xmath64 : * the connected component @xmath6 of the space of framed curves contains infinitely many components of the space of legendrian curves .
( see proposition [ interpretationconditionii ] for the homological interpretation of condition @xmath64 . )
this generalization of the result of fuchs and tabachnikov and theorem [ isomorphism ] imply the following theorem .
[ isomorphismobtained ] let @xmath58 be a contact manifold with a cooriented contact structure , and let @xmath9 be a connected component of the space of legendrian curves in @xmath4 .
let @xmath6 be the connected component of the space of framed curves that contains @xmath9 .
let @xmath53 ( resp .
@xmath54 ) be the group of @xmath5-valued order @xmath41 invariants of legendrian ( resp .
framed ) knots from @xmath9 ( resp . from @xmath6 ) .
then the groups @xmath53 and @xmath54 are canonically isomorphic , provided that @xmath6 satisfies condition @xmath64 .
now we give a homological interpretation of condition @xmath64 .
[ interpretationconditionii ]
let @xmath19 be a contact manifold with a cooriented contact structure , let @xmath65 be the euler class of the contact bundle , and let @xmath6 be a component of the space of framed curves in @xmath4
. then @xmath6 does not satisfy condition @xmath64 if and only if there exists @xmath66 such that @xmath67 and @xmath23 is realizable by a mapping @xmath11 with the property that @xmath68 is a loop free homotopic to loops realized by curves from @xmath6 . for the proof of proposition [ interpretationconditionii ]
see subsection [ proofinterpretationconditionii ] .
[ homologysphere]_some immediate corollaries of theorem [ isomorphismobtained ] and the generalization of the theorem of fuchs and tabachnikov about the isomorphism of the groups of the @xmath15-valued vassiliev invariants in the case of @xmath69 .
_ proposition [ interpretationconditionii ] implies that if the contact structure is parallelizable ( and hence the euler class of the contact bundle is zero ) then all the connected components of the space of framed curves satisfy condition @xmath64 . applying theorem [ isomorphismobtained ]
we conclude that for any abelian group @xmath5 and for every connected component of the space of legendrian curves @xmath9 and for the containing it component of the space of framed curves @xmath6 the groups @xmath53 and @xmath54 of @xmath5-valued vassiliev invariants are canonically isomorphic .
clearly the value of the euler class of the contact bundle is zero if @xmath4 is an integer homology sphere .
hence for any abelian group @xmath5 we obtain the isomorphism of the groups @xmath53 and @xmath54 of @xmath5-valued vassiliev invariants .
this generalizes the theorem of d. fuchs and s. tabachnikov @xcite saying that for the standard contact @xmath14 and for @xmath13 the quotient groups @xmath70 and @xmath71 are canonically isomorphic .
the proof of this theorem of fuchs and tabachnikov was based on the fact that for the @xmath15-valued vassiliev invariants of framed knots in @xmath14 there exists the universal vassiliev invariant constructed by t. q. t. le and j. murakami @xcite .
( for unframed knots in @xmath14 the construction of the universal vassiliev invariant is the classical result of m. kontsevich @xcite , and the invariant itself is the famous kontsevich integral . )
the existence of the universal vassiliev invariant is currently known only for a very limited collection of @xmath0-manifolds , and only for @xmath5 being @xmath15 , @xmath72 , or @xmath73 .
( andersen , mattes , reshetikhin @xcite proved its existence in the case where @xmath13 and @xmath4 is the total space of an @xmath16-bundle over a compact oriented surface @xmath2 with @xmath74 . )
thus the approach used in @xcite to show the isomorphism of the quotient groups is not applicable for almost all contact @xmath0-manifolds and abelian groups @xmath5 , and theorem [ isomorphismobtained ] appears to be a strong generalization of the result of fuchs and tabachnikov . let @xmath19 be a contact manifold with a cooriented contact structure , and let @xmath6 be a connected component of the space of framed curves in @xmath4 .
theorem [ isomorphismobtained ] implies that the group of @xmath5-valued order @xmath41 invariants of legendrian knots from a connected component @xmath75 of the space of legendrian curves does not depend on the choice of a cooriented contact structure , provided that for this choice @xmath6 satisfies condition @xmath64 . and
hence in these cases the group can not be used to distinguish cooriented contact structures on @xmath4 .
( see remark [ homologysphere ] and theorems [ atoroidal ] and [ tight ] for the list of cases when the connected components of the space of framed curves are known to satisfy condition @xmath64 . )
[ arnoldsj+]*finite order arnold s @xmath1-type invariants of wave fronts on surfaces . * a very interesting class of contact manifolds satisfying the conditions of theorem [ isomorphismobtained ] is formed by the spherical cotangent bundles @xmath3 of surfaces @xmath2 with the natural contact structure on @xmath3 ( see [ definitions ] ) .
the theory of the invariants of legendrian knots in @xmath3 is often referred to as the theory of arnold s @xcite @xmath1-type invariants of fronts on a surface @xmath2 .
the natural contact structure on @xmath3 is cooriented .
( the coorientation is induced from the coorientation of the contact elements of @xmath2 . )
one can verify that for orientable @xmath2 the standard contact structure on @xmath3 is parallelizable , and hence all the components of the space of framed curves satisfy condition @xmath64 . if @xmath2 is not orientable , then the standard cooriented contact structure on @xmath3 is not parallelizable , but one can still verify ( cf .
proposition 8.2.4 @xcite ) that every connected component of the space of framed curves satisfies condition @xmath64 .
hence for any abelian group @xmath5 and for any surface @xmath2 we obtain the canonical isomorphism of the groups of @xmath5-valued order @xmath41 invariants of legendrian and of framed knots ( from the corresponding components of the spaces of legendrian and of framed curves in @xmath3 with the standard contact structure ) .
or equivalently we get that the groups of @xmath5-valued order @xmath41 @xmath1-type invariants of fronts on @xmath2 and of @xmath5-valued order @xmath41 invariants of framed knots in @xmath3 ( from the corresponding components of the two spaces ) are canonically isomorphic .
previously it was known that for @xmath76 and @xmath77 the quotient groups @xmath70 and @xmath71 are canonically isomorphic .
the proof of this result of j. w. hill @xcite was based on the fact that for the @xmath15-valued vassiliev invariants of framed knots in @xmath78 there exists the universal vassiliev invariant constructed by v. goryunov @xcite .
( for unframed knots in @xmath14 the existence of the universal vassiliev invariant is the classical result of m. kontsevich @xcite , and the invariant itself is the famous kontsevich integral . )
our results generalize the result of j. w. hill ( even in the case of @xmath79 .
the following theorem describes another big class of contact manifolds for which the groups of vassiliev invariants of legendrian and of framed knots ( from the corresponding components of the two spaces of curves ) are canonically isomorphic .
[ atoroidal ] let @xmath19 be a contact manifold ( with a cooriented contact structure ) such that @xmath80 , and for every mapping @xmath81 of the two - torus the homomorphism @xmath82 is not injective .
then all the components of the space of framed curves in @xmath4 satisfy condition @xmath64 , and hence the groups of @xmath42-valued order @xmath41 invariants of legendrian and of framed knots ( from the corresponding components of the spaces of legendrian and framed curves ) are canonically isomorphic . for the proof of theorem [ atoroidal ] see subsection [ proofatoridal ] .
[ negativecurvature ] _ the isomorphism of the groups of vassiliev invariants in the case of closed manifolds admitting a metric of negative sectional curvature and other corollaries of theorem [ atoroidal ] .
_ let @xmath4 be a closed manifold admitting a metric of negative sectional curvature . a well - known theorem by a. preissman ( see @xcite pp .
258 - 265 ) says that every nontrivial commutative subgroup of the fundamental group of a closed @xmath0-dimensional manifold of negative sectional curvature is infinite cyclic .
hence for every mapping @xmath81 the kernel of @xmath83 is nontrivial .
it is also known that the universal covering of such @xmath4 is diffeomorphic to @xmath14 , and hence @xmath80 .
thus every closed manifold @xmath4 admitting a metric of negative sectional curvature satisfies all the conditions of theorem [ atoroidal ] and for an arbitrary cooriented contact structure on such @xmath4 we obtain the isomorphism of the groups of @xmath5-valued order @xmath41 invariants of legendrian and of framed knots from the corresponding components of the spaces of legendrian and of framed curves . another important class of contact manifolds for which every connected component of the space of framed curves satisfies condition @xmath64 is formed by contact manifolds with a tight contact structure .
the following theorem appeared as a result of discussions of stefan nemirovski and the author .
[ tight ] let @xmath19 be a contact manifold with a tight cooriented contact structure
. then all the components of the space of framed curves in @xmath4 satisfy condition @xmath64 , and hence the groups of @xmath42-valued order @xmath41 invariants of legendrian and of framed knots ( from the corresponding components of the spaces of legendrian and framed curves ) are canonically isomorphic .
for the proof of theorem [ tight ] see subsection [ prooftight ] .
let @xmath4 be a contact manifold with a parallelized contact structure @xmath28 .
let @xmath84 be a connected component of the space of transverse curves in @xmath19 , and let @xmath6 be the connected component of the space of framed curves that contains @xmath84 .
( such a component exists because a transverse curve in a manifold with a parallelized contact structure is naturally framed , and a path in the space of transverse curves corresponds to a path in the space of framed curves . )
let @xmath85 ( resp .
@xmath54 ) be the group of @xmath5-valued order @xmath41 invariants of transverse ( resp .
framed ) knots from @xmath84 ( resp . from @xmath6 ) .
clearly every invariant @xmath55 restricted to the category of transverse knots in @xmath84 is an element @xmath86 .
this gives a homomorphism @xmath87 .
[ isomorphismtransverse ] let @xmath58 be a contact manifold with a parallelized contact structure .
let @xmath84 be a connected component of the space of transverse curves in @xmath19 , and let @xmath6 be the component of the space of framed curves that contains @xmath84 . then the following two statements * a * and * b * are equivalent . a : : @xmath59 for any @xmath88 and any knots @xmath89 representing isotopic framed knots .
b : : @xmath90 is a canonical isomorphism .
the proof of theorem [ isomorphismtransverse ] is analogous to the proof of theorem [ isomorphism ] .
similar to the case of theorem [ isomorphism ] the proof of theorem [ isomorphismtransverse ] becomes obvious if the mapping from the isotopy classes of transverse knots in @xmath84 to the isotopy classes of framed knots in @xmath6 is surjective .
however in general this mapping is not surjective and to obtain the proof of theorem [ isomorphismtransverse ] one follows the ideas of the proof of theorem [ isomorphism ] ( the famous bennequin inequality shows that this mapping is not surjective even for the standard contact @xmath14 . ) thus to obtain the isomorphism between the groups @xmath54 and @xmath85 it suffices to show that statement @xmath63 of theorem [ isomorphismtransverse ] is true for the connected components @xmath84 and @xmath6 of the spaces of transverse and of framed curves . in @xcite d. fuchs and s.
tabachnikov showed that statement @xmath63 holds for all the connected components of the space of transverse curves in the case where @xmath4 is the standard contact @xmath14 and @xmath13 .
( one can verify that the proof of this theorem of fuchs and tabachnikov goes through for @xmath5 being any abelian group . )
they later observed @xcite that since their proof of this fact is mostly local , the similar fact should be true for a big class of contact manifolds .
however in fact the proof of their theorem is not completely local and is based on the existence of a well - defined bennequin invariant for a transverse knot in @xmath14 . unfortunately the bennequin invariant is not well - defined unless the knot is zero homologous . and
the generalization of this theorem to the case of manifolds other than @xmath14 meets certain difficulties that are similar to the ones we meet when we generalize the analogous theorem of fuchs and tabachnikov for legendrian knots , see [ reasonstoappear ] .
we imitate the arguments we use in [ reasonstoappear ] and obtain that statement * a * of theorem [ isomorphismtransverse ] is true for any contact @xmath0-manifold with a parallelized contact structure .
( observe that in the case of transverse knots , on the contrary to the case of legendrian knots , no extra conditions on the contact manifold appear to be needed for the statement * a * to be true . )
thus we get the following theorem .
[ isomorphismobtainedtransverse ] let @xmath58 be a contact manifold with a parallelized contact structure , and let @xmath84 be a connected component of the space of transverse curves in @xmath4 .
let @xmath6 be the connected component of the space of framed curves that contains @xmath84 .
let @xmath85 ( resp .
@xmath54 ) be the group of @xmath5-valued order @xmath41 invariants of transverse ( resp .
framed ) knots from @xmath84 ( resp . from @xmath6 ) .
then the groups @xmath85 and @xmath54 are canonically isomorphic .
this generalizes the theorem of d. fuchs and s. tabachnikov @xcite saying that for the standard contact @xmath14 and for @xmath13 the quotient groups @xmath91 and @xmath71 are canonically isomorphic .
the proof of this theorem of fuchs and tabachnikov was based on the fact that for the @xmath15-valued vassiliev invariants of framed knots in @xmath14 there exists the universal vassiliev invariant constructed by t. q. t. le and j. murakami @xcite .
( for unframed knots in @xmath14 the construction of the universal vassiliev invariant is the classical result of m. kontsevich @xcite , and the invariant itself is the well - known kontsevich integral . )
the existence of the universal vassiliev invariant is currently known only for a very limited collection of @xmath0-manifolds , and only for @xmath5 being @xmath15 , @xmath72 or @xmath73 .
( andersen , mattes , reshetikhin @xcite proved its existence in the case where @xmath13 and @xmath4 is the total space of an @xmath16-bundle over a compact oriented surface @xmath2 with @xmath74 . )
thus the approach used in @xcite to show the isomorphism of the quotient groups is not applicable for almost all contact @xmath0-manifolds and abelian groups @xmath5 , and theorem [ isomorphismobtainedtransverse ] appears to be a strong generalization of the result of fuchs and tabachnikov . let @xmath19 be a contact manifold with a parallelized contact structure , and let @xmath6 be a connected component of the space of framed curves in @xmath4 .
theorem [ isomorphismobtainedtransverse ] implies that for any @xmath92 the group of @xmath5-valued order @xmath41 invariants of transverse knots from a connected component of the space of transverse curves contained in @xmath6 does not depend on the choice of a parallelized contact structure .
hence this group can not be used to distinguish parallelized contact structures on @xmath4 .
in this section we construct a big class of examples when vassiliev invariants distinguish legendrian knots that realize isotopic framed knots and are homotopic as legendrian curves .
theorem [ isomorphism ] says that in these examples the groups of vassiliev invariants of legendrian and of framed knots are not canonically isomorphic , and we obtain the first known examples when these groups are not canonically isomorphic .
theorem of r. lutz @xcite says that for an arbitrary orientable @xmath0-manifold @xmath4 every homotopy class of distributions of @xmath21-planes tangent to @xmath4 contains a contact structure .
( the theorem of ya . eliashberg @xcite says even more that every homotopy class of the distributions of @xmath21-planes tangent to @xmath4 contains a positive overtwisted contact structure . )
however in our constructions we will use only the euler classes of contact bundles .
for this reason we start with the following proposition .
[ existcontact ] let @xmath4 be an oriented @xmath0-manifold and let @xmath93 be an element of @xmath12
. then @xmath94 can be realized as the euler class of a cooriented contact structure on @xmath4 if and only if @xmath95 , for some @xmath96 . for the proof of proposition [ existcontact ]
see subsection [ proofexistcontact ] .
let @xmath28 be a cooriented contact structure on @xmath97 such that the euler class of the contact bundle is nonzero .
( the euler class takes values in @xmath98 , and proposition [ existcontact ] says that for any even @xmath99 there exists a cooriented contact structure on @xmath17 with the euler class @xmath100 . )
let @xmath34 be a knot in @xmath17 that crosses exactly once one of the spheres @xmath101 .
the theorem of chow @xcite and rashevskii @xcite says that there exists a legendrian knot @xmath102 that is @xmath103-small isotopic to @xmath34 as an unframed knot .
let @xmath104 be the legendrian knot that is the same as @xmath102 everywhere except of a small piece located in a chart contactomorphic to the standard contact @xmath14 where it is changed as it is shown in figure [ change.fig ] ( see [ description ] ) .
[ example1 ] a : : legendrian knots @xmath102 and @xmath104 belong to the same component of the space of legendrian curves and realize isotopic framed knots .
b : : there exists a @xmath18-valued order one invariant @xmath105 of legendrian knots , such that @xmath106 . for the proof of theorem [ example1 ] see subsection [ proofexample1 ] .
let @xmath107 , @xmath108 , be the knot that is the same as @xmath102 everywhere except of a small piece located in a chart contactomorphic to the standard contact @xmath14 where it is changed in the way described by the addition of @xmath100 zigzags shown in figure [ change.fig ] .
the proof of theorem [ example1 ] implies that all @xmath107 s are homotopic as legendrian curves and realize isotopic framed knots , but for all @xmath109 legendrian knots @xmath110 and @xmath111 are not legendrian isotopic .
the order one invariant of legendrian knots @xmath105 constructed in the proof of theorem [ example1 ] has the property that @xmath112 .
hence this @xmath105 distinguishes all the @xmath107 s .
[ overtwisted]*examples of nonisotopic legendrian knots with overtwisted complements that realize isotopic framed knots and are homotopic as legendrian immersions .
* let @xmath113 be an embedded into @xmath4 disk centered at a point @xmath114 .
the theorem of eliashberg @xcite says that every homotopy class of distributions of @xmath21-planes tangent to @xmath4 contains an overtwisted contact structure that has @xmath113 as the standard overtwisted disk . in the example of theorem [ example1 ]
we can start with an overtwisted contact structure that has @xmath113 as an overtwisted disk and with an unframed knot @xmath34 that is far away from @xmath113 . then since both @xmath102 and @xmath104 were constructed using a @xmath103-small approximation of @xmath34
, we can assume that they are also far away from @xmath113 .
and we have constructed examples of nonisotopic legendrian knots with overtwisted complements that realize isotopic framed knots and are homotopic as legendrian immersions .
previously such examples were unknown and the theorem of ya .
eliashberg and m. fraser @xcite says that such examples are impossible if the ambient manifold is @xmath115 .
[ otherexamples ] _ below we describe another big family of examples where finite order invariants distinguish legendrian knots that realize isotopic framed knots and are homotopic as legendrian immersions .
_ let @xmath2 be a nonorientable surface that can be decomposed as a connected sum of the klein bottle @xmath34 and a surface @xmath116 .
let @xmath4 be an orientable manifold that admits a structure of a locally trivial @xmath7-fibration @xmath117 .
( for example one can take @xmath4 to be the spherical tangent bundle @xmath118 of @xmath2 . )
consider an @xmath7-fibration @xmath119 induced from @xmath120 by the mapping @xmath121 that corresponds to the solid loop in figure [ example2.fig ] .
( in this figure the enumeration of the end points of the arcs indicates which pairs of points should be identified to obtain the loop . )
since the solid loop is an orientation preserving loop in @xmath2 , we get that @xmath122 ( torus ) .
put @xmath123 to be the natural mapping of the total space of the induced fibration @xmath119 into the total space of @xmath117 .
a homology class in @xmath124 projecting to the dashed loop in figure [ example2.fig ] has intersection @xmath22 with the class @xmath125\in h_2(m , \z)$ ] realized by @xmath126 .
thus there exists @xmath127 such that @xmath128)=1 $ ] .
proposition [ existcontact ] says that for every @xmath129 the class @xmath130 is realizable as the euler class of a cooriented contact structure on @xmath4 .
thus for every @xmath129 there exists a cooriented contact structure on @xmath4 such that the value of the euler class of the contact bundle on @xmath125 $ ] is equal to @xmath131
. let @xmath28 be a cooriented contact structure on @xmath4 such that the euler class @xmath132 of the contact bundle satisfies @xmath133)=2r$ ] , for some nonzero @xmath129 .
let @xmath34 be an arbitrary legendrian knot such that its projection to @xmath2 ( considered as a loop ) is free homotopic to the solid loop in figure [ example2.fig ] .
let @xmath134 be legendrian knots that are the same as @xmath34 everywhere except of a chart ( contactomorphic to the standard contact @xmath14 ) where @xmath104 and @xmath135 are different from @xmath34 as it is described in figure [ example4.fig ] , see [ description ] .
( the number of cusps in figure [ example4.fig ] is @xmath133)=2r\neq 0 $ ] . )
[ example2 ] the knots @xmath104 and @xmath135 described above belong to the same component @xmath9 of the space of legendrian curves and realize isotopic framed knots . there exists a @xmath18-valued order one invariant @xmath105 of legendrian knots from @xmath9 such that @xmath136 . for the proof of theorem [ example2 ] see subsection [ proofexample2 ] . similarly to [ overtwisted ] one verifies that the contact structure and the knots @xmath104 and @xmath135 in the statement of theorem [ example2 ] can be chosen so that the restrictions of the contact structure to the complements of @xmath104 and of @xmath135 are overtwisted . using the ideas of the proof of theorem [ example2 ]
one can construct many other examples of legendrian knots that can be distinguished by vassiliev invariants of legendrian knots even though they realize isotopic framed knots and are homotopic as legendrian immersions .
for example as a solid loop in figure [ example2.fig ] we could take any loop @xmath137 such that the number of double points that separate @xmath137 into two orientation reversing loops is odd , and the value of the euler class of the contact bundle on @xmath125\in h_2(m,\z)$ ] is nonzero .
[ commute ] let @xmath138 be a locally trivial @xmath7-fibration of an oriented manifold @xmath139 over a ( not necessarily orientable ) manifold @xmath140 .
let @xmath141 be the class of an oriented @xmath7-fiber of @xmath120 , and let @xmath23 be an element of @xmath142 . then : _ proof of proposition [ commute ] .
_ if we move an oriented fiber along the loop @xmath146 , then in the end it comes to itself either with the same or with the opposite orientation .
it is easy to see that it comes to itself with the opposite orientation if and only if @xmath144 is an orientation reversing loop in @xmath140 .
_ proof of proposition [ preissman ] .
_ it is well - known that any closed @xmath2 , other than @xmath150 admits a hyperbolic metric .
( it is induced from the universal covering of @xmath2 by the hyperbolic plane . ) the theorem of a. preissman ( see @xcite pp . 258 - 265 ) says that if @xmath4 is a closed riemannian manifold of negative sectional curvature , then any nontrivial abelian subgroup @xmath151 is isomorphic to @xmath18 .
thus if @xmath152 is closed , then any nontrivial commutative @xmath153 is infinite cyclic . if @xmath2 is not closed , then the statement of the proposition is also true because in this case @xmath2 is homotopy equivalent to a bouquet of circles .
[ toughandtechnical ] let @xmath154 ( klein bottle ) be a surface not necessarily closed or orientable .
let @xmath4 be an orientable @xmath0-manifold , and let @xmath117 be a locally trivial @xmath7-fibration .
let @xmath155 be the class of an oriented @xmath7-fiber of @xmath120 , and let @xmath156 be an element with @xmath157 .
let @xmath137 be an element of the centralizer @xmath158 of @xmath23 .
then there exist @xmath159 and nonzero @xmath160 such that @xmath161 .
[ prooftoughandtechnical ] _ proof of proposition [ toughandtechnical ] . _
since @xmath23 and @xmath137 commute in @xmath162 we get that @xmath163 and @xmath164 commute in @xmath149 . proposition [ preissman ] and the fact that @xmath157 imply that there exist @xmath165 with @xmath166 , @xmath99 , and nonzero @xmath160 such that @xmath167 and @xmath168 .
_ an important homomorphism._[hansen ] let @xmath139 be a manifold , let @xmath171 be the space of free loops in @xmath139 , and let @xmath172 be a loop .
an element @xmath173 is realizable by a mapping @xmath174 with @xmath175 .
let @xmath176 be the element corresponding to the trace of the point @xmath177 under the homotopy of @xmath178 described by @xmath23 .
let @xmath179 be the homomorphism that maps @xmath180 to @xmath181 .
since the @xmath21-cell of @xmath182 is glued to the @xmath22-skeleton along the commutation relation of the meridian and of the longitude of @xmath182 , we get that @xmath179 is a surjective homomorphism of @xmath183 onto the centralizer @xmath184 of @xmath185 .
if @xmath186 for @xmath187 , then the mappings @xmath188 and @xmath189 of @xmath182 corresponding to these loops can be deformed to be identical on the @xmath22-skeleton of @xmath182 . clearly the obstruction for @xmath188 and @xmath189 to be homotopic as mappings of @xmath182 ( with the mapping of the @xmath22-skeleton of @xmath182 fixed under homotopy ) is an element of @xmath190 obtained by gluing together the boundaries of the @xmath21-cells of the two tori .
in particular we get the proposition of v. l. hansen @xcite saying that @xmath191 is an isomorphism , provided that @xmath192 .
[ h - principleforcurves ] _
@xmath43-principle for curves in @xmath4 .
_ for a @xmath0-dimensional manifold @xmath4 we put @xmath193 to be the manifold obtained by the fiberwise spherization of the tangent bundle of @xmath4 , and we put @xmath194 to be the corresponding locally trivial @xmath195-fibration .
the @xmath43-principle ( that can be found in @xcite ) says that the space of curves in @xmath4 is weak homotopy equivalent to @xmath196 ( the space of free loops in @xmath193 ) .
the weak homotopy equivalence is given by mapping a curve @xmath34 to a loop @xmath197 that sends a point @xmath198 to the point of @xmath193 corresponding to the direction of the velocity vector of @xmath34 at @xmath199 .
[ obstruction ] let @xmath104 and @xmath135 be two framed knots that coincide pointwise as embeddings of @xmath7 .
then there is an integer obstruction @xmath200 for them to be isotopic as framed knots with the embeddings of @xmath7 fixed under the isotopy .
this obstruction is calculated as follows .
let @xmath201 be the knot obtained by shifting @xmath104 along the framing and reversing the orientation on the shifted copy .
together @xmath104 and @xmath201 bound a thin strip .
we put @xmath202 to be the intersection number of the strip with a very small shift of @xmath135 along its framing .
for two singular framed knots @xmath207 and @xmath208 with @xmath32 transverse double points that coincide pointwise as immersions of @xmath7 , we put @xmath209 to be the value of @xmath210 on the nonsingular framed knots @xmath104 and @xmath135 that coincide pointwise as embeddings of @xmath7 and are obtained from @xmath207 and @xmath208 by resolving each pair of the corresponding double points of @xmath207 and of @xmath208 in the same way .
( the value @xmath211 does not depend on the resolution as soon as the corresponding double points of the two knots are resolved in exactly the same way . ) as before @xmath211 is the integer valued obstruction for @xmath207 and @xmath208 to be isotopic as singular framed knots with the immersion of @xmath7 corresponding to the two knots fixed under isotopy .
for a singular framed knot @xmath212 with @xmath32 transverse double points we denote by @xmath213 , @xmath99 , the isotopy class of a singular framed knot with @xmath32 transverse double points that coincides with @xmath212 as an immersion of @xmath7 and has @xmath214 .
[ homotopy ] let @xmath104 and @xmath135 be framed knots ( resp . singular framed knots with @xmath32 transverse double points ) that coincide pointwise as embeddings ( resp .
immersions ) of @xmath7
. then @xmath104 and @xmath135 are homotopic as framed knots ( resp .
singular framed knots with @xmath32 transverse double points ) if and only if @xmath202 is even .
[ proofhomotopy]_proof of proposition [ homotopy ] .
_ clearly if @xmath202 is even , then @xmath104 and @xmath135 are framed homotopic .
( we can change the obstruction by two by creating a small kink and passing through a double point at its vertex . )
every oriented @xmath0-dimensional manifold @xmath4 is parallelizable , and hence it admits a @xmath215-structure .
a framed curve @xmath34 in @xmath4 represents a loop in the principal @xmath216-bundle of @xmath217 .
( the @xmath0-frame corresponding to a point of @xmath34 is the velocity vector , the framing vector , and the unique third vector of unit length such that the @xmath0-frame defines the positive orientation of @xmath4 . )
one observes that the values of the @xmath215-structure on the loops in the principal @xmath216-bundle of @xmath217 realized by @xmath104 and @xmath135 are different provided that @xmath202 is odd .
but these values do not change under homotopy of framed curves . hence
if @xmath202 is odd , then @xmath104 and @xmath135 are not framed homotopic .
[ strange ] using the self - linking invariant of framed knots one can easily show that if @xmath104 and @xmath135 in [ obstruction ] are pointwise coinciding zero - homologous framed knots and @xmath218 , then @xmath104 is not isotopic to @xmath135 in the category of framed knots .
however for knots that are not zero - homologous this is not generally true , see [ proofaexample1 ] .
for this reason we introduce the following definitions .
if for an unframed knot @xmath34 there exist isotopic framed knots @xmath104 and @xmath135 that coincide with @xmath34 pointwise and have @xmath218 , then we say that @xmath34 _ admits finitely many framings_. for @xmath34 that admits finitely many framings we put _ the number of framings @xmath219 of _ @xmath34 to be the minimal positive integer @xmath220 such that there exist isotopic framed knots @xmath104 and @xmath135 that coincide with @xmath34 pointwise and have @xmath221 .
one can easily show that if @xmath34 admits finitely many framings , then there are exactly @xmath219 isotopy classes of framed knots realizing the isotopy class of the unframed knot @xmath34 .
proposition [ homotopy ] implies that @xmath219 is even .
[ decrease ] let @xmath19 be a contact @xmath0-manifold with a cooriented contact structure , let @xmath6 be a connected component of the space of framed curves , and let @xmath8 be a connected component of the space of legendrian curves in @xmath58 .
a : : let @xmath34 be an unframed knot obtained by forgetting the framing on a knot from @xmath6 . then there exists a legendrian knot from @xmath222 realizing the isotopy class of @xmath34 .
b : : if @xmath203 is an isotopy class of framed knots in @xmath6 that is realizable by a legendrian knot from @xmath9 , then the isotopy class of @xmath223 ( see [ obstruction ] ) is also realizable by a legendrian knot from @xmath9 .
c : : let @xmath40 be an unframed singular knot with @xmath32 double points obtained by forgetting the framing on a singular knot from @xmath6 . then there exists a singular legendrian knot from @xmath224 realizing the isotopy class of @xmath40 .
d : : if @xmath225 is an isotopy class of singular framed knots in @xmath6 that is realizable by a singular legendrian knot from @xmath9 , then the isotopy class of @xmath226 is also realizable by a singular legendrian knot from @xmath9 .
[ proofdecrease ] _ proof of statement * a * of proposition [ decrease ] .
_ let @xmath44 be the fiberwise spherization of the @xmath21-dimensional contact vector bundle , and let @xmath45 be the corresponding locally trivial @xmath7-fibration .
we denote by @xmath227 the class of an oriented @xmath7-fiber of @xmath228 . for a legendrian curve @xmath229 denote by @xmath230 the loop in @xmath44 obtained by mapping a point @xmath198 to the point of @xmath44 corresponding to the direction of the velocity vector of @xmath31 at @xmath231 .
the @xmath43-principle [ h - principlelegendrian ] says that legendrian curves @xmath104 and @xmath135 in @xmath4 belong to the same component of the space of legendrian curves in @xmath4 if and only if @xmath232 and @xmath233 are free homotopic loops in @xmath44 .
w. l. chow @xcite and p. k. rashevskii @xcite showed that every unframed knot @xmath34 is isotopic to a legendrian knot @xmath31 ( and this isotopy can be made @xmath103-small ) .
deforming @xmath34 we can assume ( see [ h - principleforcurves ] ) that : * 1 : * @xmath34 and @xmath31 coincide in the neighborhood of @xmath234 , * 2 : * @xmath34 and @xmath31 realize the same element @xmath235\in \pi_1(m , k_l(1))$ ] , and * 3 : * that liftings to @xmath44 of legendrian curves from @xmath9 are free homotopic to a loop @xmath23 in @xmath44 such that @xmath236 and @xmath237 \in \pi_1(m , k_l(1))$ ] .
take a chart of @xmath4 ( that is contactomorphic to the standard contact @xmath14 ) containing a piece of the legendrian knot . from the formula for the maslov number deduced in
@xcite it is easy to see that the modifications of the legendrian knot corresponding to the insertions of two cusps shown in figure [ twocusp.fig ] ( see [ description ] ) induce multiplication by @xmath241 of the lifting of @xmath31 to an element of @xmath242 .
( here the sign depends on the choice of an orientation of the fiber used to define @xmath238 . ) performing this operation sufficiently many times we obtain the legendrian knot from @xmath9 realizing the isotopy class of the unframed knot @xmath34 .
_ proof of statement * b * of proposition [ decrease ] .
_ take a chart of @xmath4 ( that is contactomorphic to the standard contact @xmath14 ) containing a piece of the knot @xmath203 and perform the homotopy in @xmath9 shown in figure [ kink.fig ] , see [ description ] .
( observe that a self - tangency point of the projection of a legendrian curve in @xmath14 to the @xmath48-plane corresponds to a double point of the legendrian curve . ) straightforward verification ( cf .
the formula for the bennequin invariant deduced in @xcite ) shows that the legendrian knot we obtain in the end of the homotopy realizes @xmath223 .
the fact that statement * b * of theorem [ isomorphism ] implies statement * a * is clear .
thus we have to show that statement * a * implies statement .
this is done by showing that there exists a homomorphism @xmath243 such that @xmath244 and @xmath245 .
[ definitionofpsi]_definition of @xmath249 . _
if the isotopy class of the knot @xmath250 is realizable by a legendrian knot @xmath251 , then put @xmath252 .
the value @xmath253 is well - defined because if @xmath254 is another knot realizing @xmath34 , then @xmath255 by statement * a * of theorem [ isomorphism ] .
let @xmath256 be the component of the space of unframed curves that corresponds to forgetting framings on the curves from @xmath6 .
propositions [ decrease ] and [ homotopy ] imply that if an unframed knot @xmath257 admits finitely many framings ( see [ strange ] ) , then all the isotopy classes of framed knots from @xmath6 realizing the isotopy class of the unframed knot @xmath258 are realizable by legendrian knots from @xmath9 .
thus we have defined the value of @xmath249 on all the framed knots from @xmath6 that realize unframed knots admitting finitely many framings .
if @xmath257 admits infinitely many framings , then either * 1 ) * all the isotopy classes of framed knots from @xmath6 realizing the isotopy class of @xmath258 are realizable by legendrian knots from @xmath9 or * 2 ) * there exists a knot @xmath259 realizing the isotopy class of @xmath258 such that @xmath203 is realizable by a legendrian knot from @xmath9 and @xmath260 ( see [ obstruction ] ) is not realizable by a legendrian knot from @xmath9 .
( in this case @xmath261 etc . also are not realizable by legendrian knots from @xmath9 , see [ decrease ] . ) in the case * 1 ) * the value of @xmath249 is already defined on all the framed knots from @xmath6 realizing @xmath258 . in the case * 2 ) * put @xmath262 ( proposition [ decrease ] implies that the sum on the right hand side is well - defined . ) similarly put @xmath263 now we have defined @xmath249 on all the framed knots ( from @xmath6 ) realizing @xmath258 . doing this for all @xmath258 for which case
* 2 ) * holds we define the value of @xmath249 on all the knots from @xmath6 .
if @xmath260 is realizable by a legendrian knot @xmath265 , then consider a singular legendrian knot @xmath266 with @xmath35 double points that are vertices of @xmath35 small kinks such that we get @xmath31 if we resolve all the double points positively staying in the class of the legendrian knots .
( to create @xmath266 we perform the first half of the homotopy shown in figure [ kink.fig ] in @xmath267 places on @xmath265 . )
let @xmath268 be the set of the @xmath36 possible resolutions of the double points of @xmath266
. for @xmath269 put @xmath270 to be the sign of the resolution , and put @xmath271 to be the nonsingular legendrian knot obtained via the resolution @xmath272 .
since @xmath39 is an order @xmath41 invariant of legendrian knots we get that @xmath273 ( observe that if we resolve @xmath100 double points of @xmath266 negatively , then we get the isotopy class of @xmath274 . ) this finishes the proof of the proposition .
let @xmath275 be a singular framed knot with @xmath35 double points .
let @xmath268 be the set of the @xmath36 possible resolutions of the double points of @xmath40
. for @xmath276 put @xmath270 to be the sign of the resolution , and put @xmath277 to be the isotopy class of the knot obtained via the resolution @xmath272 .
if the isotopy class of @xmath40 is realizable by a singular legendrian knot from @xmath9 , then identity holds for @xmath40 , since @xmath39 is an order @xmath41 invariant of legendrian knots ( and the value of @xmath249 on a framed knot @xmath248 realizable by a legendrian knot @xmath251 was put to be @xmath280 ) .
if @xmath281 admits finitely many framings , then all the isotopy classes of singular framed knots from @xmath6 realizing the isotopy class of @xmath281 are realizable by singular legendrian knots from @xmath9 , and we get that identity holds for @xmath40 .
if @xmath281 admits infinitely many framings and all the isotopy classes of singular framed knots from @xmath6 realizing @xmath281 are realizable by singular legendrian knots from @xmath9 , then automatically holds for @xmath40 . if @xmath281 admits infinitely many framings but not all the isotopy classes of singular framed knots from @xmath6 realizing @xmath281 are realizable by singular legendrian knots from @xmath9
then put @xmath282 to be the framed knot realizing @xmath281 that is realizable by a singular legendrian knot from @xmath9 and such that @xmath283 , @xmath284 , are not realizable by singular legendrian knots from @xmath9 .
proposition [ decrease ] says that @xmath285 , @xmath284 , are realizable by singular legendrian knots from @xmath9 and hence identity holds for @xmath285 , @xmath286 . using proposition [ mainidentity ] and the fact that identity holds for @xmath285 , @xmath286 , we show that holds for @xmath287 .
namely , @xmath288 considering the values of @xmath55 on the @xmath36 possible resolutions of a singular framed knot with @xmath267 singular small kinks we get that @xmath51 should satisfy identity .
hence @xmath245 and this finishes the proof of theorem [ isomorphism ] .
the proof of the theorem of fuchs and tabachnikov that says that statement @xmath63 of theorem [ isomorphism ] is true for all the connected components of the space of legendrian curves when the ambient contact manifold is the standard contact @xmath14 is based on the following three observations : 1 : : there are two types of cusps arising under the projection of the part of a legendrian knot that is contained in a darboux chart to the @xmath48-plane ( see [ description ] ) .
they are formed by cusps for which the branch of the projection of the knot going away from the cusp is locally located respectively above or below the tangent line at the cusp point , see figure [ twocusp.fig ] . for a legendrian knot @xmath34 and @xmath290
we denote by @xmath291 the legendrian knot obtained from @xmath34 by the modification corresponding to an addition of @xmath100 cusp pairs of the first type and @xmath292 cusp pairs of the second type to the projection of the part of @xmath34 located in a darboux chart .
+ let @xmath104 and @xmath135 be legendrian knots in the standard contact @xmath14 that realize isotopic unframed knots . then for any @xmath293 and @xmath294 large enough there exist @xmath295 such that the legendrian knot @xmath296 is legendrian isotopic to @xmath297 .
2 : : if there exists @xmath92 such that legendrian knots @xmath298 and @xmath299 are legendrian isotopic , then every vassiliev invariant of legendrian knots takes equal values on @xmath104 and on @xmath135 . 3 : : the number @xmath32 from the previous observation exists if the ambient contact manifold is @xmath14 and the legendrian knots @xmath104 and @xmath135 belong to the same component of the space of legendrian curves and realize isotopic framed knots .
the first two observation are true for any contact @xmath0-manifold ( since the proof of the corresponding facts is local ) .
but the number @xmath32 from the statement of the third observation does not exist in general . in the case of the ambient manifold being @xmath14 fuchs and
tabachnikov showed the existence of such @xmath32 using the explicit calculation involving the maslov classes and bennequin invariants of legendrian knots . however in order for the bennequin invariant to be well - defined the knots have to be zero - homologous , and in order for the maslov class to be well - defined the knots have to be zero - homologous or the contact structure has to be parallelizable .
below we show that such @xmath32 exists for any @xmath104 and @xmath135 that realize isotopic framed knots and belong to the same component of the space of legendrian curves , provided that the connected component @xmath6 of the space of framed curves that contains @xmath104 and @xmath135 satisfies condition @xmath64 .
( we assume that the contact structure on @xmath4 is cooriented . ) _ proof of the fact that @xmath302 can be chosen so that @xmath303 .
_ let @xmath305\rightarrow m$ ] be the isotopy changing @xmath104 to @xmath135 in the category of framed knots . analyzing the proof of fuchs and tabachnikov one verifies that for @xmath306 large enough
the legendrian isotopy @xmath307 changing @xmath301 to @xmath308 can be chosen so that for every @xmath309 $ ] the legendrian knot @xmath310 is contained in a thin tubular neighborhood @xmath311 of @xmath312 and is isotopic ( as an unframed knot ) to @xmath313 inside @xmath311 . for two framed knots @xmath313 and @xmath314 realizing unframed knots that are isotopic inside @xmath311 there is a well - defined @xmath18-valued obstruction to be isotopic inside @xmath311 in the category of framed knots .
this obstruction is the difference of the self - linking numbers of the inclusions of @xmath313 and @xmath315 into @xmath14 induced by an identification of @xmath311 with the standard solid torus in @xmath14 .
( one verifies that for @xmath313 and @xmath315 that are isotopic as unframed knots inside @xmath311 this difference does not depend on the choice of the identification of @xmath311 with the standard solid torus in @xmath14 . ) from the formula for the bennequin invariant stated in @xcite one gets that the value of the obstruction for @xmath301 to be isotopic as a framed knot to @xmath104 inside @xmath316 is equal to @xmath317 .
similarly the value of the obstruction for @xmath308 to be isotopic as a framed knot to @xmath135 inside @xmath318 is equal to @xmath319 .
clearly the value of the obstruction for @xmath320 to be isotopic to @xmath321 inside @xmath311 does not depend on @xmath33 ( for the isotopy @xmath322 changing @xmath104 to @xmath135 in the category of framed knots ) , and we get that @xmath303 . _
proof of the fact that if @xmath6 satisfies condition @xmath64 , then @xmath304 .
_ let @xmath323 be the class of an oriented @xmath7-fiber of @xmath45 . from the @xmath43-principles for legendrian and for unframed curves ( see [ h - principlelegendrian ] and [ h - principleforcurves ] )
one obtains that every component of the space of legendrian curves contained in @xmath6 corresponds to the conjugacy class of @xmath324 , for some @xmath325 .
( connected components of the space of free loops in @xmath44 are naturally identified with the conjugacy classes of the elements of @xmath47 . ) from the formula for the maslov number deduced in @xcite and the @xmath43-principle for legendrian curves one gets that @xmath296 is contained in the component of the space of legendrian curves that corresponds to the conjugacy class of @xmath327 .
using the fact that @xmath232 and @xmath233 are conjugate in @xmath47 ( since @xmath104 and @xmath135 are legendrian homotopic ) and the fact that since the contact structure is cooriented @xmath238 is in the center of @xmath47 ( see [ commute ] ) , we get that @xmath297 is contained in the component that corresponds to the conjugacy class of @xmath328 . since @xmath296 and @xmath297 are legendrian isotopic ( and hence legendrian homotopic ) we get that @xmath329 and @xmath330 are conjugate in @xmath47 , and using [ commute ] we get that @xmath232 is conjugate to @xmath331 .
but since @xmath6 satisfies condition @xmath64 we have @xmath332 , and hence @xmath304 . from the identities @xmath303 and @xmath304 one gets that @xmath333 and @xmath334 .
assume that @xmath335 .
( the case where @xmath336 is treated similarly . ) put @xmath337 .
it is easy to show that since @xmath301 and @xmath297 are legendrian isotopic , then @xmath338 and @xmath339 are also legendrian isotopic .
( basically one can keep the @xmath340 extra cusp pairs close together on a small piece of the projection of the part of the knot contained in a darboux chart during the whole isotopy process . ) but @xmath338 and @xmath339 are obtained from @xmath104 and @xmath135 by the modification corresponding to the addition of @xmath341 pairs of cusps of each of the two types , and we can take @xmath32 from the observation * 2 * to be @xmath341 .
this shows that @xmath104 and @xmath135 can not be distinguished by the vassiliev invariants of legendrian knots provided that @xmath6 satisfies condition @xmath64 , and that @xmath104 and @xmath135 realize isotopic framed knots and are homotopic as legendrian immersions .
hence statement * a * of theorem [ isomorphism ] is true provided that @xmath6 satisfies condition @xmath64 .
the @xmath43-principle for curves [ h - principleforcurves ] says that the set @xmath256 of the connected components of the space of curves in @xmath4 is naturally identified with the set of the connected components of the space of free loops in the spherical tangent bundle @xmath193 of @xmath4 .
hence it is also naturally identified with the set of conjugacy classes of the elements of @xmath342 .
( from the long homotopy sequence of the fibration @xmath194 we see that it is also naturally identified with the set of conjugacy classes of the elements of @xmath162 . ) choose a @xmath215-structure on @xmath4 .
it is easy to see ( cf .
[ homotopy ] and [ proofhomotopy ] ) that the set @xmath343 of the connected components of the space of framed curves in @xmath4 is identified with the product @xmath344 . here
the @xmath345-factor is the value of the @xmath215-structure on the loop in the principal @xmath216-bundle of @xmath217 that corresponds to a framed curve from the connected component , see [ proofhomotopy ] .
( this value does not depend on the choice of the framed curve in the component . )
the @xmath43-principle for the legendrian curves says that the set of the connected components of the space of legendrian curves is naturally identified with the set of homotopy classes of free loops in @xmath44 ( the spherical contact bundle of @xmath4 ) .
hence it is also naturally identified with the set of conjugacy classes of the elements of @xmath47 .
since every contact manifold is oriented and the contact structure was assumed to be cooriented , we get that the planes of the contact structure are naturally oriented .
this orientation induces the orientation of the @xmath7-fibers of @xmath45 .
put @xmath323 to be the class of the oriented @xmath7-fiber of @xmath45 .
the theorem of chow @xcite and rashevskii @xcite says that every connected component of the space of curves contains a legendrian curve .
straightforward verification shows that the insertion of the zig - zag into the legendrian curve @xmath34 ( see figure [ twocusp.fig ] ) changes the value of the @xmath215-structure on the corresponding framed curve .
it is easy to verify ( see @xcite ) that the two connected components of the space of legendrian curves that contain respectively @xmath34 and @xmath34 with the extra zigzag correspond to the conjugacy classes of @xmath346 and of @xmath347 ( or of @xmath348 ) in @xmath47 .
( we obtain @xmath347 or @xmath348 depending on which of the two possible zig - zags we insert . )
let @xmath8 be a connected component of the space of legendrian curves in @xmath58 that corresponds to the conjugacy class of @xmath349 .
then every connected component @xmath350 of the space of legendrian curves corresponds to the conjugacy class of @xmath351 , for some @xmath160 .
put @xmath93 to be the euler class of the contact bundle .
consider the locally - trivial @xmath7-fibration @xmath362 induced by @xmath322 from the @xmath7-fibration @xmath363 .
one can verify that @xmath364 is the euler class of @xmath120 . on the other hand
the euler class of @xmath120 is @xmath365 and it is naturally identified with the value of @xmath93 on the homology class realized by @xmath126 .
this implies that if @xmath6 does not satisfy condition @xmath64 , then there exists a homology class @xmath23 from the statement of the proposition . on the other hand
the existence of the class @xmath23 from the statement of the proposition implies that there exists a legendrian curve @xmath366 such that @xmath346 is conjugate to @xmath367 , for @xmath32 being the value of @xmath93 ( the euler class of the contact bundle ) on the homology class realized by @xmath126 .
( proposition [ existcontact ] says that @xmath95 , for some @xmath368 , and hence @xmath32 is even . )
this means that @xmath6 does not satisfy condition @xmath64 and we have proved proposition [ interpretationconditionii ] .
similar to [ proofinterpretationconditionii ] we get that to prove that all the components of the space of framed curves satisfy condition @xmath64 , it suffices to show that @xmath346 and @xmath353 are not conjugate in @xmath47 , for all @xmath369 and @xmath370 .
let @xmath371 and @xmath160 be such that @xmath372 we have to show that @xmath373 .
identity implies that @xmath356 and @xmath376 commute in @xmath162 .
hence there exists a mapping of the two - torus @xmath377 such that @xmath378 and @xmath379 . by the assumption of the theorem
@xmath380 has a nontrivial kernel . thus there
exist @xmath159 with at least one of @xmath100 and @xmath292 being nonzero such that @xmath381 , and hence @xmath382 thus @xmath384 . applying to the last identity we get that @xmath385 . since @xmath80
we see that @xmath238 has infinite order in @xmath47 , and hence @xmath386 .
if @xmath32 is zero , then we are done .
hence we have to look at the case of @xmath387 .
( we assumed that @xmath383 . ) from we get that @xmath388 , and hence by proposition [ commute ] @xmath389 is in the center of @xmath47 .
thus @xmath390 . on the other hand
using we get that @xmath391 .
since @xmath238 has infinite order in @xmath47 we get that @xmath392 . by our assumptions
@xmath383 and we have @xmath373 .
this finishes the proof of theorem [ atoroidal ] .
let @xmath94 be the euler class of the contact bundle of @xmath19 .
proposition [ interpretationconditionii ] implies that it suffices to show that @xmath393 , for every homology class @xmath394 realizable by a mapping @xmath81 .
the result of d. gabai ( see corollary 6.18 @xcite ) implies that every @xmath395 realizable by a mapping of @xmath182 can be realized by a collection of spheres and of a torus that are embedded into @xmath4 .
finally the result of ya .
eliashberg ( see @xcite theorem @xmath396 ) says that for a tight contact structure the value of @xmath93 on any embedded torus or sphere is zero .
hence @xmath393 .
this finishes the proof of theorem [ tight ] .
since the contact structure is cooriented we get that the tangent bundle @xmath217 is isomorphic to the sum @xmath398 of the oriented contact bundle @xmath28 and the trivial oriented line bundle @xmath399 .
the tangent bundle of every orientable @xmath0-manifold is trivializable and we get that the second stiefel - whitney class of the contact bundle is zero . but
the second stiefel - whitney of @xmath28 is the projection of the euler class of @xmath28 under the natural mapping @xmath400 , and we get that @xmath95 for some @xmath397 .
consider an oriented @xmath21-dimensional vector bundle @xmath401 over @xmath4 with the euler class @xmath402 .
the second stiefel - whitney class @xmath403 of @xmath401 is zero , since it is the projection of @xmath404 . since @xmath401 is an oriented vector bundle we have @xmath405 .
consider the sum @xmath406 of @xmath401 with the trivial oriented @xmath22-dimensional vector bundle @xmath399 .
clearly the total stiefel - whitney class of the @xmath0-dimensional oriented vector bundle @xmath406 is equal to @xmath22 , and the euler class of @xmath406 is equal to @xmath407 . using the interpretation of the stiefel - whitney and the euler classes of @xmath406 as obstructions for the trivialization of @xmath406
, we get that @xmath406 is trivializable . since the tangent bundle of an oriented @xmath0-dimensional manifold is trivializable , we see that @xmath401 is isomorphic to an oriented sub - bundle of @xmath217 . since @xmath4 is oriented this sub - bundle of @xmath217 is also cooriented
now the theorem of lutz @xcite , that says that every homotopy class of distributions of @xmath21-planes tangent to @xmath4 contains a contact structure , implies the existence of a cooriented contact structure with the euler class @xmath93 .
[ proofaexample1 ] _ proof of statement * a * of theorem [ example1 ] . _
clearly ( see figure [ homotopy.fig ] ) the two legendrian knots @xmath102 and @xmath104 belong to the same component of the space of legendrian curves .
it is easy to see that if @xmath102 realizes the isotopy class of a framed knot @xmath408 , then @xmath104 realizes the isotopy class of @xmath409 ( see [ obstruction ] for the definition of @xmath409 ) .
below we show that @xmath408 and @xmath409 are isotopic framed knots .
let @xmath410 be the sphere that crosses @xmath408 at exactly one point , and let @xmath411\times s^2 $ ] be a thin tubular neighborhood of @xmath412 .
fix @xmath413 ( below called the north pole ) and the direction in @xmath414 ( below called the zero meridian ) .
we can assume that the knot @xmath408 inside @xmath411\times s^2 $ ] looks as follows : it intersects each @xmath415\times s^2 $ ] at the north pole of the corresponding sphere , and the framing of the knot is parallel to the zero meridian .
consider an automorphism @xmath416 that is identical outside of @xmath411\times s^2 $ ] such that it rotates each @xmath417\times s^2 $ ] by @xmath418 around the north pole in the clockwise direction . clearly under this automorphism @xmath408
gets two extra negative twists of the framing and @xmath419 . on the other hand
it is easy to see that @xmath420 is diffeotopic to the identity , since it corresponds to the contractible loop in @xmath421 .
hence we see that @xmath408 and @xmath409 are isotopic framed knots .
this finishes the proof of statement * a * of theorem [ example1 ] .
[ props1xs2 ] let @xmath28 be a cooriented contact structure on @xmath97 with a nonzero euler class @xmath93 of the contact bundle .
let @xmath44 be the spherical contact bundle , let @xmath45 be the corresponding locally trivial @xmath7-fibration , and let @xmath323 be the class of an oriented @xmath7-fiber of @xmath228 .
then @xmath238 is of finite order in @xmath47 and @xmath422 . _
proof of proposition [ props1xs2 ] .
_ consider the oriented @xmath21-plane bundle @xmath423 that is the restriction of the contact bundle over @xmath4 to the sphere @xmath424 .
the euler class of @xmath120 is the value of @xmath93 on the homology class realized by @xmath425 , and hence is nonzero .
let @xmath426 be the manifold obtained by the fiberwise spherization of @xmath120 , and let @xmath427 be the corresponding locally trivial @xmath7-fibration .
since the euler class of @xmath120 is nonzero we get that a certain multiple of the class of the fiber of @xmath428 is homologous to zero . but
@xmath429 is generated by the class of the fiber , and hence the class of the fiber of @xmath428 is of finite order in @xmath429 .
this implies that @xmath323 is of finite order .
_ proof of statement * b * of theorem [ example1 ] .
_ let @xmath9 be the connected component of the space of legendrian curves that contains @xmath102 and @xmath104 .
figure [ homotopy.fig ] shows that @xmath102 can be changed to @xmath104 ( in the space of legendrian curves ) by a sequence of isotopies and one passage through a transverse double point . hence if there exists a @xmath18-valued invariant @xmath105 of legendrian knots from @xmath9 that increases by one under every positive passage through a transverse double point of a legendrian knot , then it distinguishes @xmath102 and @xmath104 .
( clearly if such @xmath105 does exist , then it is an order one invariant of legendrian knots . ) below we show the existence of such @xmath105 in the connected component @xmath9 .
put @xmath430 .
let @xmath431 be a legendrian knot , and let @xmath432 be a generic path in @xmath9 connecting @xmath102 and @xmath433 .
let @xmath434 be the set of moments when @xmath432 crosses the discriminant ( i.e. the subspace of singular knots ) in @xmath9 , and let @xmath435 , @xmath436 , be the signs of these crossings . for a generic path
@xmath437 put @xmath438 .
it is clear that if @xmath105 ( with @xmath430 ) does exist , then @xmath439 . to show that @xmath105 does exist we have to verify that for every legendrian knot @xmath431 and for a generic path @xmath432 connecting @xmath433 to @xmath102 the value of @xmath440
does not depend on the choice of a generic path @xmath432 connecting @xmath102 and @xmath433 , or equivalently we have to show that @xmath441 for every generic closed loop @xmath432 connecting @xmath102 to itself .
there are two codimension two strata of the discriminant of @xmath9 .
they are formed respectively by singular legendrian knots with two transverse double points , and by legendrian knots with one double point at which the two intersecting branches are tangent of order one .
straightforward verification shows that @xmath442 , for every small closed loop @xmath137 going around a codimension two stratum of @xmath9 .
this implies that for every generic loop @xmath432 connecting @xmath102 to itself the value of @xmath440 depends only on the element of @xmath443 realized by @xmath432 . hence to prove the existence of @xmath105 it suffices to show that @xmath441 for every @xmath444 .
the @xmath43-principle says that the space of legendrian curves in @xmath19 is weak homotopy equivalent to the space of free loops @xmath46 in the spherical contact bundle @xmath44 of @xmath97 .
( the mapping giving the equivalence lifts a legendrian curve @xmath34 to a loop @xmath346 in @xmath44 by sending @xmath198 to the point in @xmath44 that corresponds to the direction of the velocity vector of @xmath34 at @xmath199 . )
thus @xmath443 is naturally isomorphic to @xmath448 .
proposition [ props1xs2 ] says that @xmath422 and from [ hansen ] we get that @xmath449 is isomorphic to the centralizer @xmath450 of @xmath451 . using propositions [ commute ] and [ props1xs2 ] we see that either @xmath452 or @xmath453 , for some nonzero @xmath454 .
hence there exists @xmath160 and nonzero @xmath455 such that @xmath456 .
( one should take @xmath32 and @xmath210 to be divisible by @xmath120 if @xmath453 . ) but the loop @xmath23 in @xmath443 corresponding to @xmath457 is just the sliding of @xmath102 @xmath32 times along itself according to the orientation .
( this deformation is induced by the rotation of the parameterizing circle . )
this loop does not intersect the discriminant , and hence @xmath458 .
this finishes the proof of statement * b * of theorem [ example1 ] .
[ part1]_@xmath104 and @xmath135 are homotopic legendrian curves and they realize isotopic framed knots .
_ let @xmath459 be the class of the @xmath7-fiber of the fibration @xmath45 .
the @xmath43-principle says that the connected component of the space of legendrian curves that contains @xmath34 corresponds to the conjugacy class of @xmath370 . from the formula for the maslov number deduced in @xcite
it is easy to see that the connected components containing @xmath104 and @xmath135 correspond to the conjugacy classes of @xmath460 and of @xmath461 .
let @xmath462 be an element projecting to the class @xmath155 of the @xmath7-fiber of @xmath117 .
the value of the euler class of the contact bundle on the homology class realized by @xmath126 is equal to @xmath463 .
( here @xmath322 is the mapping from the description of the euler class of the contact bundle . ) and because of the reasons explained in the proof of proposition [ interpretationconditionii ] we get that @xmath464 , for the legendrian knot @xmath34 used to construct @xmath104 and @xmath135 .
now proposition [ commute ] implies that @xmath232 and @xmath233 are conjugate in @xmath47 and hence @xmath104 and @xmath135 are in the same component of the space of legendrian curves .
the fact that @xmath104 and @xmath135 realize isotopic framed knots is clear , because as unframed knots they are the same , and as it is shown in @xcite every pair of extra cusps corresponds to the negative extra twist of the framing .
[ ideaexample2]_the idea of the proof of the fact that @xmath104 and @xmath135 can be distinguished by an order one invariant of legendrian knots .
_ let @xmath465 be a point in @xmath4 .
let @xmath40 be a singular unframed knot with one double point .
the double point separates @xmath40 into two oriented loops .
deform @xmath40 preserving the double point , so that the double point is located at @xmath465 .
choosing one of the two loops of @xmath40 we obtain an ordered set of two elements @xmath466 , or which is the same an element @xmath467 .
clearly there is a unique element of the set @xmath468 that corresponds to the original singular unframed knot @xmath40 , where @xmath468 is is the quotient set of @xmath469 modulo the consequent actions of the following groups : 1 : : @xmath162 whose element @xmath401 acts on @xmath470 by sending it to @xmath471 .
( this corresponds to the ambiguity in deforming @xmath40 , so that the double point is located at @xmath465 . ) 2 : : @xmath345 that acts via the cyclic permutation of the two summands .
( this corresponds to the ambiguity in the choice of one of the two loops of @xmath40 . )
assume that @xmath478 is an invariant of legendrian knots from @xmath9 such that under every ( generic transverse ) positive passage through a discriminant in @xmath9 it increases by @xmath479 , where @xmath40 is the unframed singular knot corresponding to the crossing of the discriminant .
clearly such @xmath478 is an order one invariant of framed knots from @xmath9 . to prove the theorem we show the existence of such @xmath478 , and
then we show that it distinguishes @xmath104 and @xmath135 .
[ existence]_the existence of @xmath478 .
_ let @xmath432 be a generic path in @xmath9 that starts with @xmath104 .
let @xmath434 be the set of instances when @xmath432 crosses the discriminant ( i.e. the subspace of singular knots ) in @xmath9 , and let @xmath435 , @xmath436 , be the signs of these crossings .
let @xmath480 be those instances for which the value of @xmath481 on the corresponding singular unframed knots is @xmath22 .
for a generic path @xmath437 put @xmath482 .
let @xmath256 be the connected component of the space of unframed curves obtained by forgetting the framings on curves from @xmath6 , and let @xmath201 be the unframed knot obtained by forgetting the framing on @xmath484 . similarly to the above for a generic path @xmath432 in @xmath256 starting with @xmath201 we put @xmath485 .
( as above @xmath486 is the set of instances when the value of @xmath481 on the singular unframed knots obtained under @xmath432 is equal to @xmath22 , and @xmath487 , @xmath488 , are the signs of the corresponding crossings of the discriminant . )
the codimension two stratum of the discriminant of @xmath256 consists of singular curves whose only singularities are two distinct transverse double points .
straightforward verification shows that @xmath489 for any small loop @xmath137 going around the codimension two stratum . hence @xmath490 is a homomorphism .
there are two codimension two strata in @xmath9 .
they consist of respectively singular legendrian curves whose only singularities are two distinct transverse double points and of singular legendrian curves whose only singularity is one double point at which the two branches are tangent . considerations similar to the ones above show that @xmath491 is a homomorphism .
the @xmath43-principle says that the space of legendrian curves in @xmath19 is weak homotopy equivalent to the space of free loops in the spherical contact bundle @xmath44 of @xmath4 .
( the mapping that gives the equivalence lifts a legendrian curve @xmath34 in @xmath58 to a loop @xmath346 in @xmath44 by mapping @xmath495 to the point of @xmath44 that corresponds to the velocity vector of @xmath34 at @xmath199 . ) since @xmath422 for @xmath4 from the statement of the theorem , we obtain ( see [ hansen ] ) the natural isomorphism @xmath496 . since @xmath497 and @xmath18 is torsion free
, we get that to show the existence of @xmath478 it suffices to show that for every @xmath498 there exist @xmath369 and @xmath499 such that @xmath500 and @xmath501 .
clearly @xmath517 and since @xmath518 is surjective there exists a loop @xmath519 such that @xmath520 .
let @xmath521 be the loop corresponding to the deformation under which @xmath104 slides once around itself according to the orientation of @xmath104 .
( this deformation is induced by the rotation of the circle parameterizing @xmath104 . ) clearly @xmath522 does not cross the discriminant and hence @xmath523 . to prove the existence of @xmath478 it suffices to show that @xmath524 , for @xmath499 such that @xmath525 .
but this @xmath432 is @xmath526 .
thus it suffices to show that @xmath527 .
since @xmath523 we get that @xmath528 and thus it suffices to show that @xmath529 . in
[ explain ] we show that @xmath530 can be realized as a power of the loop described by the deformation shown in figure [ obstruction.fig ] .
then since one of the loops of the only singular knot arising under this deformation is contractible and the value of @xmath533 on such a singular knot is zero we get that @xmath534 .
this finishes the proof of the existence of @xmath478 modulo the explanations given in [ explain ] .
[ explain ] _ now we show that @xmath530 can be realized as a sequence of loops described by the deformation shown in figure [ obstruction.fig ] . _ the @xmath43-principle for curves [ h - principleforcurves ] says that @xmath535 is weak homotopy equivalent to the space of free loops @xmath196 in the spherical tangent bundle @xmath193 of @xmath4 .
in particular @xmath536 . in subsubsection [ hansen ]
we introduced a surjective homomorphism @xmath33 from @xmath537 onto @xmath538 .
let @xmath539 be loops such that @xmath540 .
as it was explained in [ hansen ] the obstruction for @xmath23 and @xmath137 to be homotopic is an element of @xmath541 .
since every orientable @xmath0-manifold is parallelizable we get that @xmath542 .
clearly @xmath80 for @xmath4 from the statement of the theorem and hence @xmath543 .
consider the loop @xmath544 that looks the same as @xmath23 except for a small period of time when we perform the deformation shown in figure [ obstruction.fig ] .
clearly @xmath545 , and straightforward verification show that the @xmath18-valued obstruction for @xmath544 and @xmath137 to be homotopic differs by one from the obstruction for @xmath23 and @xmath137 to be homotopic . hence performing this operation ( or its inverse ) sufficiently many times we can change @xmath23 to be homotopic to @xmath137 .
[ distinguish ] _ let us show that @xmath478 distinguishes @xmath104 and @xmath135 .
_ let @xmath547\rightarrow \mathcal l$ ] be a generic path connecting @xmath104 and @xmath135 . to prove the theorem we have to show that @xmath548 .
let @xmath201 ( resp @xmath549 ) be the unframed knot obtained by forgetting the framing on @xmath104 ( resp @xmath135 ) .
let @xmath550\rightarrow \mathcal c$ ] be the isotopy that deforms @xmath201 into @xmath549 in the category of unframed curves under which @xmath201 all the time stays in a thin tubular neighborhood of @xmath549 .
consider a homotopy @xmath551 that corresponds to a product of paths @xmath552 .
( @xmath553 connects @xmath201 to itself . ) clearly @xmath554 . for a loop @xmath555 ( that connects @xmath201 to itself )
put @xmath556 to be a mapping such that for every @xmath495 the mappings @xmath557 and @xmath558 are the same .
the value of the euler class of the contact bundle on the homology class realized by @xmath559 is equal to @xmath560 .
_ loop @xmath564 . _
since @xmath565 is an orientation preserving loop and @xmath4 is orientable , we get that the @xmath7-fibration over @xmath7 ( parameterizing the knots ) induced from @xmath117 by @xmath566 is trivializable .
hence we can coherently orient the fibers of this fibration .
the orientation of the @xmath7-fiber over @xmath198 induces the orientation of the @xmath7-fiber of @xmath120 that contains @xmath567 .
the loop @xmath564 is the deformation of @xmath201 under which every point of @xmath201 slides once around the fiber of @xmath120 that contains this point ( staying inside the fiber ) in the direction specified by the orientation of the fiber corresponding to this point .
the @xmath43-principle for curves says that @xmath256 is weak homotopy equivalent to the space of free loops @xmath196 in the spherical tangent bundle @xmath193 of @xmath4 .
( the mapping that gives the equivalence lifts a curve @xmath34 in @xmath4 to a loop @xmath346 in @xmath193 by mapping @xmath495 to the point of @xmath193 that corresponds to the velocity vector of @xmath34 at @xmath199 . )
let @xmath568 be the surjective homomorphism described in [ hansen ] , and let @xmath569 be the element that projects to the class @xmath155 of the @xmath7-fiber of @xmath117 .
using proposition [ toughandtechnical ] one verifies that for every @xmath561 there exist @xmath369 such @xmath570 , for some @xmath159 .
let @xmath571 be the loop described in figure [ obstruction.fig ] .
( it is easy to see that @xmath571 is in the center of @xmath572 . )
similar to [ explain ] we get that @xmath573 for some @xmath574 .
[ bholdsforgamma ] _ if @xmath576 then @xmath577 and * b * holds for @xmath432 .
_ the loop @xmath564 crosses the discriminant twice , both crossings occur with the same sign and the values of @xmath533 on the corresponding singular knots are equal to one .
these crossings occur in the fiber over the double point of @xmath565 .
( since the double point of @xmath565 separates it into two orientation reversing loops , the two points of @xmath201 contained in this fiber induce opposite orientations of it , and the two branches of @xmath201 that intersect the fiber slide in the opposite directions under @xmath564 . )
hence @xmath578 . ( the sign depends on the orientation of the @xmath7-fibers of @xmath579 used to induce the orientations of the fibers containing the points of @xmath201 . )
* acknowledgments .
* i am very grateful to stefan nemirovski , serge tabachnikov and oleg viro for the valuable discussions and suggestions .
i am deeply thankful to h. geiges and a. stoimenow for the valuable suggestions , and to o. baues , m. bhupal , n. a campo , a. cattaneo , j. frhlich , j. latschev , a. shumakovich , and v. turaev for many valuable discussions .
this paper was written during my stay at the max - planck - institut fr mathematik ( mpim ) , bonn , and it is a continuation of the research conducted at the eth zurich @xcite .
i would like to thank the directors and the staff of the mpim and the staff of the eth for hospitality and for providing the excellent working conditions .
99999 v.i .
invarianty i perestroiki ploskih frontov , _ trudy mat .
inst . steklov .
* 209 * ( 1995 ) ; english translation : _ invariants and perestroikas of wave fronts on the plane , _ singularities of smooth mappings with additional structures , _ proc .
steklov inst .
* 209 * ( 1995 ) , pp .
j. e. andersen , j. mattes , n. reshetikhin : _ quantization of the algebra of chord diagrams , _ math .
cambridge philos .
, vol . * 124 * no . 3 ( 1998 ) ,
451467 w. l. chow : _ ber systemevon linearen partiellen differentialgleichungen erster ordnung , _ math .
* 117 * ( 1939 ) , pp .
98105 m. do carmo : _ riemannian geometry , _ birkhuser , boston ( 1992 ) ya .
eliashberg : _
contact @xmath0-manifolds twenty years since j. martinet s work , _ ann .
fourier ( grenoble ) * 42 * no . 1 - 2 ( 1992 ) , pp .
165192 ya .
eliashberg : _ classification of overtwisted contact structures on @xmath0-manifolds , _ invent .
* 98 * ( 1989 ) , pp . 623637 ya .
eliashberg , m. fraser : _ classification of topologically trivial legendrian knots , _ crm proceedings and lecture notes , vol . * 15 * ( 1998 ) , pp .
1751 d. fuchs and s. tabachnikov : _ invariants of legendrian and transverse knots in the standard contact space , _ topology vol . * 36 * , no . 5 ( 1997 ) ,
10251053 d. fuchs and s. tabachnikov : joint results , private communication with s. tabachnikov ( 1999 ) d. gabai : _ foliations and topology of @xmath0-manifolds , _ j. differential geom . * 18 * no . 3 ( 1983 ) , pp .
445503 v. goryunov : _ vassiliev type invariants in arnold s @xmath583-theory of plane curves without direct self - tangencies , _ topology vol . * 37 * no . 3 ( 1998 ) ,
m. gromov : _ partial differential relations , _ springer - verlag , berlin heidelberg ( 1986 ) v. l. hansen : _ on the fundamental group of the mapping space , _ compositio math .
* 28 * ( 1974 ) , pp .
3336 j. w. hill : _ vassiliev - type invariants in @xmath1-theory of planar fronts without dangerous self - tangencies , _ c. r. acad .
paris sr .
, vol . * 324 * no . 5 ( 1997 ) ,
537542 m. kontsevich : _
vassiliev s knot invariants , _
gelfand seminar , pp .
137150 , adv .
soviet math .
, 16 , part 2 , amer .
soc . , providence , ri ( 1993 ) .
r. lutz : _ structures de contact sur les fibrs principaux en cercles de dimension @xmath0 , _ ann .
fourier , vol . * 3 * ( 1977 ) , pp .
115 t. q. t. le and j. murakami : _ the universal vassiliev - kontsevich invariant for framed oriented links , _ compositio math .
, vol . * 102 * no . 1 ( 1996 ) pp .
k. rashevskii : _ about the possibility to connect any two points of a completely nonholonomic space by an admissible curve , _ uchen .
libknecht ped .
ser . mat .
* 2 * ( 1938 ) , pp .
8394 v. tchernov : _ arnold - type invariants of wave fronts on surfaces _ , to appear in topology ; preprint http://xxx.lanl.gov/ math.gt/9901133 ( 1999 ) v. tchernov : _ finite order invariants of legendrian , transverse and framed knots in contact @xmath0-manifolds , _
preprint http://xxx.lanl.gov math.sg/9907118 ( 1999 ) | we show that for a large class of contact @xmath0-manifolds the groups of vassiliev invariants of legendrian and of framed knots are canonically isomorphic . as a corollary
, we obtain that the group of finite order arnold s @xmath1-type invariants of wave fronts on a surface @xmath2 is isomorphic to the group of vassiliev invariants of framed knots in the spherical cotangent bundle @xmath3 of @xmath2 . on the other hand
we construct the first examples of contact manifolds for which vassiliev invariants of legendrian knots can distinguish legendrian knots that realize isotopic framed knots and are homotopic as legendrian immersions . |
the study of the surface tension of deconfined quark matter has attracted much attention recently @xcite because a detailed knowledge of it may contribute to a better comprehension of the physics of compact star interiors .
in fact , surface tension plays a crucial role in quark matter nucleation during the formation of compact stellar objects , because it determines the nucleation rate and the associated critical size of the nucleated drops @xcite .
it is also determinant in the formation of mixed phases at the core of hybrid stars which may arise only if the surface tension is smaller than a critical value of the order of tens of mev / @xmath0 @xcite .
also , surface tension affects decisively the properties of the most external layers of a strange star which may fragment into a charge - separated mixture , involving positively - charged strangelets immersed in a negatively charged sea of electrons , presumably forming a crystalline solid crust @xcite . this would happen below a critical surface tension which is typically of the order of a few mev/@xmath0 @xcite .
however , in spite of its key role for compact star physics , the surface tension is still poorly known for quark matter .
early calculations by berger and jaffe gave rather low values for the surface tension , below @xmath1 @xcite .
however , larger values within @xmath2 were used in other works about quark matter droplets in neutron stars @xcite .
more recently , values around @xmath3 have been adopted for studying the effect of quark matter nucleation on the evolution of protoneutron stars @xcite .
however , much larger values have also been obtained in the literature .
estimates given in ref .
@xcite give values in the range @xmath4 and values around @xmath5 were suggested on the basis of dimensional analysis of the minimal interface between a color - flavor locked phase and nuclear matter @xcite .
in this paper we study the surface tension and the curvature energy of three - flavour quark matter in equilibrium under weak interactions within the nambu - jona - lasinio ( njl ) model .
we include the effect of color superconductivity and describe finite size effects within the multiple reflection expansion ( mre ) framework @xcite .
we shall consider only the 2sc phase , but we shall see that our conclusions are quite general and the results for other superconducting phases can be easily foreseen within the present model .
our calculations result in large values of the surface tension which disfavor the formation of mixed phases at hybrid star cores .
the article is organized as follows : in sect .
ii we present the quark matter equations of state without finite size effects .
then , in sect .
iii we introduce the mre formalism for the finite size effects . finally , in sect .
iv we present our results and conclusions .
in the present work we start from an @xmath6 njl effective model which also includes color superconducting quark - quark interactions .
the corresponding lagrangian is given by @xmath7 \label{lagrangian } \\
& + & 2h \!\ !
\sum_{a , a'=2,5,7 } \left [ \left ( \bar \psi \ i \gamma_5 \tau_a \lambda_{a ' } \ \psi_c \right ) \left ( \bar \psi_c \ i \gamma_5 \tau_a \lambda_{a ' } \ \psi \right ) \right ] \nonumber\end{aligned}\ ] ] where @xmath8 is the current mass matrix in flavour space .
in what follows we will work in the isospin symmetric limit @xmath9 , and for simplicity we do not include flavour mixing effects .
the matrices @xmath10 and @xmath11 with @xmath12 are the gell - mann matrices corresponding to the flavour and color groups respectively , and @xmath13 . in addition , in eq .
( [ lagrangian ] ) we have used the charge conjugate spinors @xmath14 and @xmath15 , where @xmath16 is the dirac conjugate spinor and @xmath17 . the next step is obtaining the grand canonical thermodynamic potential at finite temperature @xmath18 and chemical potentials @xmath19 , where @xmath20 and @xmath21 stand for flavour and color respectively .
then , we are able to calculate the relevant thermodynamic quantities .
note that first we will show the thermodynamic potential for the bulk system , and later on we will derive the effective potential for the finite size effects .
for that purpose , starting from eq .
( [ lagrangian ] ) , it is convenient to perform a standard bosonization of the theory .
thus , we introduce the bosonic fields @xmath22 , @xmath23 and @xmath24 corresponding to the sigma and pion mesons , and scalar diquark fields , respectively ; and integrate out the quark fields . in what follows we will work within the mean field approximation ( mfa ) , in which these bosonic fields are expanded around their vacuum expectation values and the corresponding fluctuations are neglected .
since the mean field values of the pion fields vanish due to symmetry reasons , in what follows we will only consider those of the sigma and diquark fields .
then , @xmath25 and @xmath26 . regarding the diquark mean field , in the present work
, we will assume that in the density region of interest only the 2sc phase might be relevant .
moreover , due to the color symmetry , one can rotate in color space to fix @xmath27 , @xmath28 .
finally , in the framework of the matsubara and nambu - gorkov formalism we obtain the following mfa thermodynamic potential @xmath29 per unit volume ( further calculation details can be found in refs .
@xcite ) @xmath30 where @xmath31 is the cut - off of the model and @xmath32 is defined by @xmath33 \nonumber \\ & & - t \ln[1+e^{-(x+y)/t } ] , \label{omeguinha}\end{aligned}\ ] ] with @xmath34 ^ 2 + \delta^2 ) ^{1/2 } , \\
x_{8,9 } = ( [ e \pm ( \mu_{ug } + \mu_{dr})/2 ] ^2 + \delta^2)^{1/2 } \ , \\ y_1 = \mu_{ub } , \quad y_2 = \mu_{db } , \quad y_{3 } = \mu_{sr } , \\
y_{4 } = \mu_{sg } , \quad y_{5 } = \mu_{sb } , \\
y_{6,7 } = ( \mu_{ur}-\mu_{dg})/2 , \quad y_{8,9 } = ( \mu_{ug } - \mu_{dr})/2 .
\end{aligned}\ ] ] in the above expressions @xmath35 and @xmath36 , with @xmath37 .
as we are working in the isospin limit , then @xmath38 which gives @xmath39 . in principle one
has nine different quark chemical potentials , corresponding to the three quark flavours ( u , d and s ) and three quark colors ( r , g , and b ) .
nevertheless , as discussed above , with our particular election of the orientation of the gap @xmath40 in the color space , there is a residual color symmetry ( between red and green colors ) . moreover , if we require the system to be in chemical equilibrium , it can be seen that all chemical potentials are not independent from each other , as it will be discussed in next section .
the total thermodynamic potential is obtained by adding to @xmath41 the contribution of the electrons and a vacuum constant .
namely , @xmath42 where @xmath43 is the thermodynamic potential of the electrons . for them
we use the expression corresponding to a free gas of ultra - relativistic fermions @xmath44 it is important to notice that in eq .
( [ qmp ] ) we have subtracted the constant @xmath45 in order to have a vanishing pressure at vanishing temperature and chemical potentials . however , this conventional prescription is merely an arbitrary way to uniquely determine the eos of the njl model without any further assumptions @xcite . in the mit bag model for instance , the pressure in the vacuum is non - vanishing . in view of this , @xmath46 is taken as a free parameter in ref .
@xcite , having in mind that tuning this constant is an easy way to control the splitting between the chiral restoration density and the deconfinement density .
nevertheless , we shall see below that @xmath46 does nt have a direct influence on the values of the surface tension and the curvature energy .
now we are ready to introduce the effects of finite size in the thermodynamic potential . for doing
so we consider the multiple reflection expansion formalism ( see refs . @xcite and references therein ) which consists in modifying the density of states for the case of a finite spherical droplet as follows @xmath47 where the surface contribution to the density of states is @xmath48 and the curvature contribution
is given in the madsen ansatz @xcite @xmath49\ ] ] to take into account the finite quark mass contribution .
the density of states of mre for massive quarks is reduced compared with the bulk one , and for a range of small momentum becomes negative .
this non - physical negative values are removed by introducing an infrared ( ir ) cutoff in momentum space @xcite .
thus , we have to perform the following replacement in order to obtain the thermodynamic quantities @xmath50 the ir cut - off @xmath51 is the largest solution of the equation @xmath52 with respect to the momentum @xmath53 . after the above replacement ,
the full thermodynamic potential for finite size spherical droplets reads : @xmath54 multiplying on both sides of the last equation by the volume of the quark matter drop and rearranging terms we arrive to the following form for @xmath55 @xmath56 where the pressure @xmath57 , the surface tension and the curvature energy density , are defined as @xcite @xmath58 @xmath59 and @xmath60 respectively .
we are considering a spherical drop , i.e. the area is @xmath61 and the curvature is @xmath62 . as mentioned above , once we have the grand thermodynamic potential @xmath55 then we can obtain the relevant thermodynamic quantities . we can readily obtain the number density of quarks of each flavor and color @xmath63 and the number density of @xmath64 ( assumed to be massless ) @xmath65 then , the corresponding number densities of each flavor , @xmath66 , and of each color , @xmath67 , in the quark phase are given by @xmath68 and @xmath69 respectively .
the baryon number density reads @xmath70 . in order to derive the eos from the above formalism it is necessary to impose a suitable number of conditions on the variables @xmath72 and @xmath40 .
the first three of these conditions arise from the fact that the thermodynamically consistent solutions correspond to the stationary points of @xmath73 with respect to @xmath74 , @xmath75 , and @xmath40 .
then , we have @xmath76 for the remaining conditions one must specify the physical situation in which one is interested in . in this work
we are interested in the study of finite size color - superconducting droplets in @xmath71-equilibrium that may form e.g. within the mixed phase of a hybrid star .
in such a case , chemical equilibrium is maintained by weak interactions among quarks , e.g. @xmath77 , @xmath78 , @xmath79 .
here we consider the situation of no neutrino trapping .
then , the lepton number is not conserved and we have four independent conserved charges , namely the electric charge @xmath80 and the three color charges @xmath81 , @xmath82 and @xmath83 .
it is more convenient to use the linear combinations @xmath84 , @xmath85 and @xmath86 , where @xmath87 ( the total quark number density ) and @xmath88 and @xmath89 are related with color asymmetries .
thus , conserved charges @xmath90 are related to four independent chemical potentials @xmath91 such that @xmath92 .
the individual quark chemical potentials @xmath19 are given by @xmath93 \nonumber \\ & & \ ; + \mu_3 ( \lambda_3)_{cc } + \mu_8 ( \lambda_8)_{cc}.\end{aligned}\ ] ] where , as before , @xmath10 and @xmath11 are the gell - mann matrices in flavor and color space respectively . from the @xmath71-equilibrium conditions we have @xmath94 for all colors
@xmath21 . then , the electron chemical potential is @xmath95 .
finally , the rest of the conditions we need to impose for electrically and color neutral matter are : @xmath96 ( note : remember that we choose a particular orientation of the gap in the color space , which introduces the @xmath97 symmetry .
thus , we trivially satisfy @xmath98 , and then @xmath99 is automatically equal to zero . ) in summary , in the case of neutron star quark matter without neutrino trapping , for each value of @xmath100 ( or @xmath101 ) and @xmath18 one can find the values of @xmath102 and @xmath103 by solving eqs .
( [ gapeq ] ) and ( [ colden ] ) supplemented by ( [ beta ] ) .
this allows us to obtain the quark matter eos in the thermodynamic region we are interested in and we can evaluate the curvature energy and surface tension of the color superconducting droplets . it is important to remark that , in general , when we numerically solve the set of equations related with all the conditions discussed above , there might be regions for which there is more than one solution for each value of t and @xmath100 . to choose the stable solution among all of them
, we require it to be an overall minimum of the thermodynamic potential .
the set of parameters we use in the present work is the following ( those in ref .
@xcite but without t hooft interactions ) , @xmath104 = 5.5 mev , @xmath105 = 112.0 mev , @xmath31 = 602.3 mev and @xmath106 = 4.638 .
moreover , we considered the ratio @xmath107 obtained from fierz transformations of the one - gluon exchange interactions . as we impose vanishing pressure at vanishing temperature and chemical potentials for fermi momentum @xmath108 ( when @xmath109 ) @xcite , thus , @xmath110 is the same as in the bulk case .
for the set of parameters we used , we found @xmath111 mev/@xmath112 ) .
however , we emphasize that the values of the surface tension @xmath113 and the curvature energy @xmath114 are not changed by the choice of @xmath115 . as we previously mentioned , the value of @xmath51 is the largest root when solving @xmath116 with respect to @xmath53 , depending on @xmath117 and r. we find that these solutions can be fitted through the following rule , @xmath118 with @xmath119 in fm and @xmath51 in mev . the coefficients for @xmath120 mev are @xmath121 and @xmath122 . for @xmath123 mev
we have @xmath124 and @xmath125 .
in fig . [ fig1 ] we show our results for the surface tension @xmath113 and curvature energy @xmath114 of the color superconducting droplets as a function of the quark chemical potential , for two different temperatures and four different radii .
the temperatures 30 mev and 5 mev are representative , respectively , of the conditions prevailing at the beginning and at the end of the cooling / deleptonization phase of protoneutron star evolution .
we have checked that the results for @xmath126 mev are almost indistinguishable of those for zero temperature , thus , in practice they also represent old and cold neutron stars .
we show results for drops with radii ranging from very small values of 5 fm , which have a large energy cost due to surface and curvature effects , to the bulk limit of @xmath127 .
for given values of @xmath119 , @xmath18 and @xmath100 there may exist more than one solution of the equations .
if more than one solution is found , the one that minimizes the thermodynamic potential is chosen . as explained in the figure caption
, the left branches correspond to the chiral symmetry broken phase and the right curves after the discontinuity to the 2sc phase . for the curves presenting negative pressures
we have introduced a dot indicating the zero pressure point .
the part of the curve to the left of the dot corresponds to @xmath128 .
note that , as previously mentioned , we subtracted @xmath46 to have vanishing pressure at vanishing @xmath18 and @xmath100 for @xmath127 .
however , @xmath46 can be taken as a free parameter as in ref .
@xcite . in this case
, the dot in the figures can move along the curves .
however , it is clear from eqs .
( [ surfacetension ] ) and ( [ curvatureenergy ] ) that a non - standard choice for @xmath46 does nt change the numerical values of @xmath113 and @xmath114 for given @xmath119 , @xmath18 and @xmath100 .
our results show that the surface tension is in the range of @xmath129 mev/@xmath0 and the curvature energy is in the range of @xmath130 mev / fm .
the large values of the surface and curvature energies are due to the linear term in the expression for @xmath32 in eq .
( [ omeguinha ] ) , which is not present in the thermodynamic potential of e.g. the mit bag model . for a given @xmath100 ,
the surface tension is an increasing function of @xmath119 .
the curvature energy behaves differently at constant @xmath100 : for very small radii ( @xmath131 fm ) it increases with @xmath119 , but in the range from @xmath132 to @xmath133 it is a decreasing function of @xmath119 . of course , both the surface and curvature contributions to the free energy per unit volume @xmath134 tend to zero as @xmath109 , as can be checked from eq .
( [ eq17 ] ) .
the different behavior with @xmath119 for @xmath113 and @xmath114 is a consequence of the functional form of the integrand together with the r - dependence of the ir cutoff .
the integral for the surface tension has a fixed upper limit ( the ultraviolet cutoff @xmath31 ) , but the lower limit ( @xmath51 ) decreases with r. then , as @xmath119 increases , @xmath51 decreases , the area under the curve increases and as a result , the surface tension increases . for the curvature energy
the effect is different .
the integrand is positive for large @xmath53 but negative for small @xmath53 . for small @xmath119
, @xmath51 falls in the positive region of the integrand .
thus , as @xmath51 decreases , the positive area under the curve increases , and the curvature energy increases . for larger @xmath119 , @xmath51 falls in the negative regionof the integrand .
consequently , as @xmath51 decreases , the negative part of the area under the curve increases , and the curvature energy decreases . the results in fig .
[ fig1 ] show that the temperature dependence of @xmath113 and @xmath114 is very weak for the values of @xmath18 that are relevant for protoneutron stars and cold neutron stars .
note also that these results are of the same order of the obtained within the njl model for just deconfined quark matter _ out of chemical equilibrium _ under weak interactions @xcite .
the present equation of state can support a two solar mass neutron star such as the pulsars psr j1614 - 2230 @xcite and psr j0348 - 0432 @xcite if a sufficiently stiff hadronic equation of state is employed for the outer layers of the star ( see for example fig 4 of @xcite ) . adding a vector interaction to the njl model , it is possible to stiffen the eos and obtain larger stellar masses .
the vector term affects the surface tension in a non - trivial way .
the chemical potentials gain an extra term @xmath135 that shift chemical equilibrium and the pressure gains a term proportional to the square of the density .
the combined effect is difficult to estimate without a full calculation that will be addressed in future work .
according to recent work @xcite , within a geometric approach to the surface tension evaluation , the surface tension can be lowered by the presence of a repulsive vector term and for magnetized quark matter the value will be further lowered @xcite . however , within the mre formalism the behavior may be different , and deserves further study .
the large values of @xmath113 and @xmath114 have strong consequences for the physics of neutron star interiors , because the energy cost of forming quark drops within the mixed phase of hybrid stars would be very large . according to @xcite , beyond a limiting value of @xmath136 mev/@xmath0
the structure of the mixed phase becomes mechanically unstable and local charge neutrality is recovered .
therefore , our results indicate that the hadron - quark interphase within a hybrid star should be a sharp discontinuity .
the consequences of the large values of @xmath113 and @xmath114 for the triggering of the deconfinement transition in neutron and protoneutron stars have been studied in @xcite .
the main difference with the present study of the quark - hadron interphase is the condition of chemical equilibrium .
the nucleation ( deconfinement ) of the first quark matter drop that triggers the conversion of the core of a hadronic star is driven by strong interactions and consequently it happens out of chemical equilibrium under weak interactions . as shown in ref .
@xcite the nucleation of quark matter is possible during the protoneutron star phase even for large values of the surface tension , because large drops ( with a size of hundreds of fm ) may have a huge nucleation rate .
these large drops are charge neutral because flavor is conserved during the deconfinement transition @xcite and therefore they can be considerably larger than the debye screening length @xmath137 of the stellar plasma which is typically @xmath138 fm . since these drops can be very large ( @xmath139 fm @xcite ) , surface and curvature effects tend to vanish .
this is not the case for droplets of quark matter in the hypothetical mixed phase of a hybrid star .
since they are electrically charged their size can not exceed @xmath140 and therefore surface and curvature have a significant energy cost , inhibiting the formation of the mixed phase .
r. balian and c. bloch , annals of physics 60 , 401 ( 1970 ) j. madsen , phys .
d * 50 * , 3328 ( 1994 ) .
o. kiriyama and a. hosaka , phys . rev .
d * 67 * , 085010 ( 2003 ) .
o. kiriyama , phys .
d * 72 * , 054009 ( 2005 ) . | in this paper we study the surface tension and the curvature energy of three - flavor quark matter in equilibrium under weak interactions within the nambu - jona - lasinio model .
we include the effect of color superconductivity and describe finite size effects within the multiple reflection expansion ( mre ) framework .
our calculations result in large values of the surface tension which disfavor the formation of mixed phases at the hadron - quark inter - phase inside a hybrid star . |
as it is well - known , the dynamics of a small josephson tunnel junction is characterized by two energies : the josephson coupling energy @xmath2 and the charging energy @xmath3 @xcite .
the behavior of the junction is determined by the competition between these two energy scales . if the josephson coupling energy dominates , a superconducting state with a well - defined phase difference @xmath4 across the junction is possible .
the junction will carry a cooper pair current @xmath5 in the absence of an external voltage @xmath1 . here
@xmath6 is the josephson critical current ( @xmath7 is the electron charge ) .
if on the other hand the charging energy dominates , an insulating state with a well - defined charge @xmath8 on the junction is possible .
this gives rise to a gap in the @xmath0-@xmath1 characteristics of the junction , associated to coulomb blockade @xcite : the cooper pair current @xmath9 up to a voltage @xmath10 .
here we defined @xmath11 , where @xmath12 is the capacitance of the junction .
the above statements are true under appropriate conditions on the impedance @xmath13 of the circuit connected to the junction , as we will discuss in some detail below . to be specific
, we will consider two configurations . the voltage - biased set - up , in fig .
[ setup]a , consists of an ideal voltage source ( voltage @xmath14 ) , connected to a series arrangement of a josephson junction and a resistance @xmath15 . in a current - biased set - up , see fig .
[ setup]b , an ideal current source ( bias current @xmath16 ) is connected to a parallel arrangement of a josephson junction and a shunt resistance @xmath15 .
the two set - ups are equivalent if we choose @xmath17 @xcite . the resistance @xmath15 plays a crucial role and determines the detailed form of the @xmath0-@xmath1 characteristics .
in particular , the way in which the transition from superconducting to insulating behavior of the junction manifests itself in the current - voltage characteristic depends not only on the parameters @xmath2 and @xmath3 , but also on @xmath15 .
the voltage biased set - up , fig .
[ setup]a , has been analyzed long ago by ivanchenko and zilberman @xcite in the limit of strong charging effects , @xmath18 .
more recently , a perturbative approach has been developed to obtain the current - voltage characteristic @xcite .
this approach is valid for finite , but small , values of @xmath19 .
specifically , in the zero temperature limit @xmath20 , a change of the position of the supercurrent peak is found upon increasing the ratio @xmath21 , where @xmath22 . for @xmath23 ,
the supercurrent peak is centered at zero bias voltage ; its width increases gradually as @xmath21 is increased .
if @xmath24 , the supercurrent peak is centered around @xmath25 .
this corresponds to the transition from superconducting behavior found for small values of the external resistance to a complete coulomb blockade of cooper pair tunneling in the presence of a resistive environment .
the range of applicability of the perturbative analysis depends on @xmath2 , @xmath3 , and @xmath15 . in the low resistance limit @xmath23 , the approach breaks down at low energy ( bias voltage or temperature ) for any value of @xmath2 . in the opposite limit @xmath26 , the condition @xmath27 needs to be satisfied .
a non - perturbative analysis was developed in @xcite , however its validity is limited to low energies , @xmath28 .
the approach discussed so far is clearly appropriate if one is interested in a small junction with @xmath29 , coupled to an environment which is not too resistive .
an example is the recent experiment by steinbach _
et al . _
@xcite , in good agreement with the theoretical predictions of refs .
however , the perturbative approach is not applicable in the case of a josephson junction with @xmath30 , coupled to a strongly resistive environment , @xmath31 , usually referred to as the underdamped limit @xcite .
nevertheless this is a relevant situation from an experimental point of view as well .
recently , watanabe and haviland measured the @xmath0-@xmath1 characteristics of a small josephson junction coupled to a tunable environment @xcite . in these experiments , @xmath32 and the environment
could be tuned to large values of the resistance such that @xmath33 @xmath34 .
no evidence of a supercurrent peak was observed ; rather a so - called bloch nose @xcite was found for small values of the current and the bias voltage .
this corresponds to an on - set of coulomb blockade at low voltage and small current , followed by subsequent back - bending of the @xmath0-@xmath1 curve to smaller values of the voltage at higher currents which indicates a cross - over to superconducting behavior .
the shape of the bloch nose depended on the resistance @xmath15 as well as on temperature .
the experimental findings could not be reconciled quantitatively with existing theories of bloch nose @xcite .
the main problem is that those theories are essentially based on the presence of dissipation due to quasiparticles . in the experiment
@xcite , the tunable environment is a long array of superconducting junctions in the insulating regime @xcite , and quasiparticles are practically absent . in this paper
we analyze the finite temperature behavior of a josephson junction with @xmath35 , coupled to a very resistive environment , in the absence of quasiparticles .
we obtain , to our knowledge for the first time , explicit formul for the current - voltage characteristics at finite temperature for a current - biased configuration with large @xmath15 , see fig .
[ setup]b .
we follow the approach pioneered in ref .
@xcite , which is based on the periodicity of the josephson coupling energy as a function of the phase difference @xmath4 across the junction . as a result of this periodicity ,
the eigenstates of the junction are of the bloch type , and the energy spectrum consists of bands of width of the order @xmath3 , separated by gaps of the order @xmath2 . in the limit @xmath35 , the gaps are large and the junction stays in the lowest bloch band at low temperature @xmath36 , as long as the bias current @xmath37 .
if @xmath38 , bloch oscillations @xcite of the voltage @xmath1 occur , such that its time - averaged value vanishes : the junction is in the superconducting state .
if @xmath15 is finite , relaxation within the lowest band modifies the dynamics of the junction .
this results in the appearance of the bloch nose : if @xmath16 is so small that relaxation prevents bloch oscillations to occur , a finite voltage state develops with no current passing through the junction ( coulomb blockade ) , all current passes through the resistance .
if @xmath16 is increased above a threshold value , bloch oscillations develop and the a superconducting zero - voltage state is reached .
our approach is valid for a large temperature range and our results can be compared quantitatively with the experimental data of ref . @xcite .
the circuit of fig .
[ setup]b can be modelled by the hamiltonian @xmath39 . here , @xmath40 describes the josephson junction in the absence of the external circuit . the charge @xmath41 and the phase @xmath42 are canonically conjugate operators , @xmath43 = 2ie$ ] .
since the hamiltonian ( [ h0 ] ) is periodic in @xmath4 , the energy spectrum consists of bands @xmath44 where @xmath45 is the band index and @xmath46 the quasimomentum associated to @xmath4 , _
i.e. _ , the quasicharge .
the spectrum @xmath44 is a periodic function of @xmath46 with period @xmath47 .
we will limit ourselves to the first brillouin zone , @xmath48 .
the corresponding eigenstates @xmath49 are of the bloch type . in the limit of interest here , @xmath50 ,
the bands are narrow compared to the gaps . throughout this paper
we will assume that all relevant energies are smaller than @xmath2 and restrict ourselves to the lowest energy band , @xmath51 , with a dispersion @xmath52 , where the bandwidth is given by @xmath53^{3/4}e^{-\sqrt{8e_j / e_c } } .
\label{u0}\ ] ] the term @xmath54 couples @xmath42 to the total current @xmath55 , which contains the constant bias current @xmath16 and a fluctuating part @xmath56 .
the fluctuations are induced by the environment ( the resistance @xmath15 in fig .
[ setup]b ) and the dynamics of @xmath57 is governed by @xmath58 . in the absence of any current , @xmath59 and quasicharge @xmath46
is a well defined variable . in the limit of small bias current @xmath16 and current fluctuations @xmath60 , the effect of the coupling term ( [ hcoupl ] ) on @xmath46
can be analyzed perturbatively . as it was shown in refs .
@xcite , a quantum langevin equation for the quasicharge operator @xmath61 can be obtained , which reads @xmath62 here , @xmath16 plays the role of a driving force and @xmath57 of a random force .
the statistical properties of the latter are determined by the hamiltonian @xmath58 .
specifically , for a resistive environment at equilibrium we have @xcite @xmath63_+ \rangle _ { \mathrm{env},\omega } = 2 ( \hbar \omega / r ) \coth [ \hbar \omega/(2k_bt ) ] \label{stat},\ ] ] where @xmath64_+$ ] denotes the anticommutator and @xmath65 an average with respect to @xmath58 . finally , the derivative @xmath66 is the voltage operator @xmath67 across the junction , the corresponding term @xmath68 acts as a damping term
. we will be particularly interested in the limit of small current fluctuations . in this limit
, quasicharge remains a well defined quantity and we can replace operators by classical variables in eq .
( [ quantlang ] ) . in the next section
, we will analyze the resulting classical langevin equation in some detail , using well - known methods for stochastic equations @xcite .
we will see under which condition current fluctuations can be considered small and we will obtain the detailed , analytical form of @xmath0-@xmath1 characteristics , as a function of the resistance @xmath15 and of temperature @xmath69 .
in the absence of fluctuations , @xmath70 , the langevin equation ( [ quantlang ] ) has a stationary solution when @xmath71 , with @xmath72 . here , @xmath73 , with @xmath74 the maximum voltage across the junction corresponding to the maximum slope of @xmath75 .
the corresponding @xmath0-@xmath1 characteristic is linear , @xmath76 . in other words ,
all the current passes through the resistance @xmath15 , no current flows through the junction ( coulomb blockade ) and the phase @xmath4 is completely undefined .
this is consistent with the fact that we treat @xmath46 as a well - defined variable .
note that the stationary solution exists only as long as @xmath77 ; for larger current , @xmath78 , the quasicharge becomes time - dependent .
we will discuss that case in sec .
[ fokker - planck ] below .
let us now take into account small fluctuations @xmath56 such that the induced charge fluctuations @xmath79 around the stationary solution @xmath80 are small .
we consider the linear part of the average current - voltage characteristics and calculate @xmath81 here the symbol @xmath82 means averaging over the fluctuations of charge . in order to find the distribution of these fluctuations ,
we write @xmath83 , such that the linearized , classical langevin equation takes the form @xmath84 where we introduced the relaxation time @xmath85 such that @xmath86 .
the time @xmath85 defines the time scale characterizing the junction dynamics .
note that this is a relatively long time compared to the @xmath87 time of the circuit : for @xmath88 , we find @xmath89 as @xmath90 is exponentially small , see eq . ( [ u0 ] )
. in order to solve the stochastic equation ( [ linlang ] ) , we should specify the correlation function @xmath91 , see eq .
( [ stat ] ) .
the frequencies @xmath92 of interest here are small , @xmath93 , determined by the long time scale @xmath85 characterizing the dynamics of quasicharge . for finite temperatures @xmath94 , the correlation function @xmath95 , independent of frequency and thus @xmath96 .
this means that the relevant current fluctuations are classical and completely uncorrelated on the long time scale @xmath85 characterizing the quasicharge dynamics .
the stochastic equation ( [ linlang ] ) can now be solved , using standard methods developed for the analysis of brownian motion @xcite . as a result ,
one obtains the probability distribution @xmath97 to find the fluctuation @xmath79 at time @xmath98 given that @xmath99 at time @xmath100 . in the long time limit , @xmath101 , this probability distribution does not depend on time and is independent of @xmath102 .
it is given by the gaussian distribution @xmath103 where we introduced the parameter @xmath104 .
using eq .
( [ gauss ] ) we immediately obtain the average square of the charge fluctuations , @xmath105 .
quasicharge can be considered well - defined as long as these fluctuations are small , @xmath106 .
this condition gives an additional restriction on temperature , @xmath107 .
together with the condition @xmath94 found above we obtain the temperature window @xmath108 this means in particular that @xmath24 for the analysis presented here to be correct . if this condition is verified , quasicharge is a well - defined variable , and quantum fluctuations can be neglected . for small @xmath16
the energy scale @xmath109 is of the order of @xmath110 , thus the condition ( [ tint ] ) reads @xmath111 .
this has a transparent physical interpretation : the temperature has to be much smaller than @xmath112 in order not to smear the bloch nose , but it must be larger than @xmath113 in order to justify the neglect of quantum fluctuations .
the experiment of ref .
@xcite was performed at very large values of @xmath21 , such that the condition ( [ tint ] ) was verified . with the help of eqs .
( [ iv ] ) and ( [ gauss ] ) we readily obtain the average voltage for the linear part of the average @xmath0-@xmath1 characteristic , @xmath114.\ ] ] equation ( [ result1 ] ) represents the main result of this section . the first term on the right hand side of eq .
( [ result1 ] ) coincides with the linear @xmath0-@xmath1 curve which was obtained above without taking into account current fluctuations .
the second term is entirely due to the thermal fluctuations , which suppress the resistance . in other words ,
the coulomb blockade is smeared due to thermal activation . in this subsection
we have considered the limit of small enough bias current , @xmath115 , such that the langevin equation could be linearized around a well - defined , stationary solution . for larger values of @xmath16 ,
no stationary solution exists : quasicharge is an oscillating function of time and the full non - linearity of the langevin equation should be taken into account . in section [ fokker - planck ] we will use a fokker - planck approach to obtain the complete current - voltage characteristics .
if the applied bias current @xmath16 exceeds @xmath116 , quasicharge is a dynamical variable even in the absence of fluctuations .
the corresponding non - linear equation of motion can be integrated , though , and @xmath117 can be found . as a result ,
the voltage @xmath1 becomes an oscillating function of time , with a period given by @xmath85 .
its time - averaged value @xmath118 over one period of the oscillation is given by @xcite @xmath119 .
\label{v0}\ ] ] in other words , on measurement time scales much longer than the period @xmath85 , the voltage oscillations average out and a stationary situation is reached . in order to investigate the influence of small current fluctuations @xmath60 on the @xmath0-@xmath1 characteristics , the full non - linear stochastic problem ( [ quantlang ] ) needs to be solved .
this can be conveniently done using a fokker - planck approach @xcite . in this approach ,
the probability distribution @xmath120 to find quasicharge @xmath46 at time @xmath98 is found , as a solution of the fokker - planck equation @xcite @xmath121w \right ) + \frac{\gamma}{2}\frac{\partial^2 w}{\partial q^2}.\ ] ] in writing this equation , we limited ourselves again to the case of classical fluctuations only , @xmath122 .
in addition , we note that this equation can be used only as long as the fluctuations are @xmath123-correlated on the time scale defined by @xmath85 .
hence , we need to restrict ourselves again to the temperature interval ( [ tint ] ) .
we are particularly interested in the stationary solution of ( [ fokker ] ) , @xmath124 , reached in the long time limit @xmath125 .
the average voltage is then given by @xmath126 where @xmath124 is normalized , @xmath127 .
in the long time limit @xmath128 , this corresponds to the measured , time - averaged voltage , @xmath129 .
let us first analyze the solution of the problem in the absence of charge fluctuations [ parameter @xmath130 equals to zero in eq .
( [ fokker ] ) ] . solving eq .
( [ fokker ] ) and using the normalization condition for the distribution function we obtain @xmath131 } , & i_x >
i_b , \end{array } \right.\end{aligned}\ ] ] where @xmath132 is the stationary solution found in section [ langevin ] . substituting eq .
( [ p ] ) into eq .
( [ v ] ) for the current - voltage characteristics we immediately obtain the linear result @xmath133 for @xmath134 ; for the opposite case @xmath78 we reproduce the result ( [ v0 ] ) . the resulting @xmath0-@xmath1 characteristic , a so - called bloch nose , is shown in fig .
[ ivcurve ] as the curve corresponding to the parameter @xmath135 .
-@xmath1 characteristics for an underdamped junction .
left panel : bias current @xmath136 as a function of voltage @xmath137 across the junction , for various values of the parameter @xmath138 .
curves from right to left correspond to @xmath139 0 , 0.004 , 0.02 , and 0.1 .
right panel : current @xmath140 through the junction as a function of @xmath141 .
curves from right to left correspond to the same values of @xmath138.,title="fig:",scaledwidth=45.0%]-@xmath1 characteristics for an underdamped junction .
left panel : bias current @xmath136 as a function of voltage @xmath137 across the junction , for various values of the parameter @xmath138 .
curves from right to left correspond to @xmath139 0 , 0.004 , 0.02 , and 0.1 .
right panel : current @xmath140 through the junction as a function of @xmath141 .
curves from right to left correspond to the same values of @xmath138.,title="fig:",scaledwidth=45.0% ] next we consider the current - voltage characteristics in the presence of fluctuations . to find the distribution function , @xmath124
, we use a fourier representation @xmath142 substituting the right hand side of eq .
( [ fourie ] ) into eq .
( [ fokker ] ) and using the periodicity condition , @xmath143 , we obtain the following recurrence relation for the distribution function @xmath144 here we introduced the short hand notation @xmath145 for @xmath146 . to find the distribution function @xmath147 in eq .
( [ recurent ] ) we use the analogy with the recurrence relation for the modified bessel function @xcite @xmath148 $ ] . here
@xmath149 is the index and @xmath150 is the argument of the modified bessel function .
a comparison with ( [ recurent ] ) yields the distribution function @xmath151 , where @xmath152 is the normalization constant . substituting this result into eq .
( [ v ] ) and using the fact that the distribution function is real , @xmath153 , we obtain , upon integration over @xmath46 , the following result for the @xmath0-@xmath1 characteristics @xmath154 equation ( [ mainresult ] ) is the main result of this section . it describes the complete current - voltage characteristics in the presence of small fluctuations .
the @xmath0-@xmath1 characteristics of eq .
( [ mainresult ] ) are shown in the left panel of fig .
[ ivcurve ] for various values of the parameter @xmath138 .
for comparison we plot the current @xmath0 through the junction as a function of @xmath155 in the right panel of fig .
[ ivcurve ] , for the same values of @xmath138 .
this would correspond to a measurement in the voltage - biased set - up , see fig .
[ setup]a .
we see that the @xmath0-@xmath1 curve corresponding to the limit without fluctuations is asymptotically reached by the finite temperature @xmath0-@xmath1 curves , both for @xmath156 and for @xmath157 .
mathematically , this is a direct result of the asymptotics of the modified bessel functions @xcite .
physically , this can be easily understood from the fact that the point @xmath158 is unstable , separating a stationary quasicharge solution from an oscillating one . in the vicinity of @xmath116 ,
any small current fluctuation will drastically change the nature of the dynamics .
thus the @xmath0-@xmath1 characteristic is strongly affected by fluctuations for @xmath159 .
we finally note that the analytical form of the result ( [ mainresult ] ) resembles the well - known result @xcite for the @xmath0-@xmath1 characteristic of an overdamped junction with @xmath160 and @xmath161 .
in fact , these two cases can be related by duality arguments . here , we studied the temperature driven diffusion of quasicharge @xmath46 , governed by eq .
( [ fokker ] ) , and found the time - averaged _ voltage _ across the junction as the average @xmath162 over the fluctuations of @xmath46 .
in an overdamped junction , it is the dual phase variable @xmath4 that diffuses , governed by an equation similar to ( [ fokker ] ) and one is interested in the time averaged _ current _ @xmath163 over the fluctuations of @xmath4 .
this is why the resulting @xmath0-@xmath1 curves can be related to each other , essentially by exchanging the role of @xmath0 and @xmath1 .
in this paper we have considered the influence of thermal fluctuations on the current - voltage characteristics of an underdamped josephson junction with @xmath164 and @xmath35 , in the absence of quasiparticles . to obtain analytical results for the @xmath0-@xmath1 characteristics
, we have analyzed the appropriate langevin equation using a fokker - planck approach .
the resulting @xmath0-@xmath1 characteristic is essentially a bloch nose @xcite , smeared by thermal fluctuations .
our results are valid as long as the inequality ( [ tint ] ) is satisfied . since we restricted our analysis to the lowest energy band we need to impose in addition @xmath165 and @xmath37 .
the present work is motivated by recent experiments @xcite on a small , voltage biased josephson junction coupled to a tunable environment .
this environment consisted of a long array with a large number ( @xmath166 ) josephson junctions .
each junction is a small squid - loop , threaded by a magnetic flux @xmath167 .
thus the josephson coupling energy @xmath168 of the array is flux - dependent and the ratio @xmath169 could be changed during the experiment .
when @xmath167 is tuned close to @xmath170 , where @xmath171 is the superconducting flux quantum , the ratio @xmath172 and the array is in the insulating regime .
we ignore quasiparticles : in order for them to reach the small josephson junction , they have to tunnel through the entire array ; the corresponding probability is negligibly small .
the frequency - dependent impedance @xmath13 of the array , as seen by the small junction , is then entirely due to the dynamics of cooper pairs . in the stationary limit of interest here
, we need to know the zero - frequency component @xmath173 .
calculations by efetov @xcite in the limit @xmath172 show that the real part @xmath15 of @xmath174 is strictly infinite at zero temperature ; at finite @xmath69 , thermal activation gives rise to a finite , large value of @xmath15 , which depends on the ratio @xmath169 , and thus on @xmath167 .
note that any residual quasiparticle contribution would be insensitive to the josephson coupling and hence independent of @xmath167 .
once @xmath175 and the parameters of the small junction are known , our results can be compared in principle with the experimental data of @xcite . + * acknowledgments * + we thank d. averin , k. efetov , d. haviland , e. mishchenko , and m. watanabe for valuable discussions .
one of us ( i. b. ) thanks laboratoire de physique et modlisation des milieux condenss for hospitality .
our research was sponsored by institut universitaire de france and cnrs - atip , grants dmr-9984002 and bsf-9800338 , as well as by the a.p .
sloan and the packard foundations .
99 m. tinkham , _ introduction to superconductivity _
( mcgraw - hill , new - york , 1996 ) .
devoret and h. grabert , _ single charge tunneling _
( plenum , new - york , 1992 ) . g.l .
ingold and yu.v .
nazarov , in ref .
m. ivanchenko and l.a .
zilberman , zh .
fiz . * 55 * , 2395 ( 1968 ) [ sov .
jetp * 28 * , 1272 ( 1969 ) ] .
ingold , h. grabert and u. eberhardt , phys .
b * 50 * , 395 ( 1994 ) . g.l .
ingold and h. grabert , phys .
* 83 * , 3721 ( 1999 ) .
a. steinbach _
et al . _ ,
lett . * 87 * , 137003 - 1 ( 2001 ) .
m. watanabe , and d. b. haviland , phys .
lett . * 86 * , 5120 ( 2001 ) ; m. watanabe and d.b .
haviland cond - mat/0301467 ( unpublished ) .
averin , a.b .
zorin , and k.k .
likharev , zh .
* 88 * , 692 ( 1985 ) [ sov .
jetp * 61*,407 ( 1985 ) ] ; k. k. likharev and a. b. zorin , j. low temp .
phys.*59 * , 347 ( 1985 ) .
g. schn and a. d. zaikin , phys . rep . * 198 * , 237 ( 1990 ) .
r. fazio and h. van der zant , phys .
rep . * 355 * , 235 ( 2001 ) .
n.g . van kampen , _
stochastic processes in physics and chemistry _
( north - holland , amsterdam , 1992 ) .
l. d. landau and e. m. lifshitz , _ statistical physics _
, part 2 , ( pergamon press , oxford , 1999 ) .
i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series , and products _
( academic press , new york , 2000 ) .
likharev , _ dynamics of josephson junctions and circuits _ ( gordon and breach , philadelphia , 1986 ) .
efetov , zh .
. fiz . * 78 * , 2017 ( 1980 ) [ sov .
jetp * 51 * , 1015 ( 1980 ) ] . | inspired by a recent experiment , we study the influence of thermal fluctuations on the @xmath0-@xmath1 characteristics of a josephson junction , coupled to a strongly resistive environment .
we obtain analytical results in the limit where the josephson energy is larger than the charging energy and quasiparticles are absent . |
to personalize an experience or make an application more secure and less accessible to undesired people , we need to be able to distinguish a person from everyone else .
it is done using marks from the users to identify them and block unauthorized access , or personalize it based on the trusted identity .
to do so , many alternatives are on the table , such as keys , passwords and cards .
the most secure options so far , however , are biometric features which can not be imitated by any other than the desired person himself .
they are divided into behavioral features that the person can uniquely create or express , such as signatures , walking rhythm , and physiological characteristics that the person possesses , such as fingerprints and iris pattern .
many works revolve around identification and verification of such data including , but not limited to , fingerprints @xcite , palmprints @xcite-@xcite , faces @xcite and iris patterns @xcite .
iris recognition systems are widely used for security applications , since they contain a rich set of features and do not change significantly over time .
they are also virtually impossible to fake .
one of the first modern algorithms for iris recognition was developed by john daugman and used 2d gabor wavelet transform @xcite . in a more recent work ,
kumar @xcite proposed to use a combination of log - gabor , haar wavelet , dct and fft based features to achieve high accuracy . in @xcite , farouk proposed an scheme which uses elastic graph matching and gabor wavelet .
each iris is represented as a labeled graph and a similarity function is defined to compare the two graphs . in @xcite , belcher used region - based sift descriptor for iris recognition and achieved a relatively good performance .
pillai @xcite proposed a unified framework based on random projections and sparse representations to achieve robust and accurate iris matching .
the reader is referred to @xcite for a comprehensive survey of iris recognition . in most of iris recognition works ,
the iris region is first detected and the iris is mapped to a rectangular region in polar coordinate .
various iris segmentation algorithms are developed during the past few years @xcite .
foreground segmentation approaches can also be used for iris segmentation @xcite , @xcite .
it is worth mentioning that no segmentation is performed to extract iris region from the eye image in our work , which makes it much easier to implement . in this work ,
two sets of features are extracted from iris images , one of them being the recently introduced set of scattering - transform features and the other one being that of textural features to capture the texture information of irises .
we believe that if these features are combined , they will provide a high discriminating power to conduct the recognition task .
after the features are extracted , their dimensionality is reduced by applying pca and then minimum distance classifier is used to recognize new iris images . skipping the segmentation step
makes our algorithm very fast and it can be easily implemented in electronic devices for real time applications using energy - efficient implementation and power management @xcite .
this algorithm is tested on the well - known iit delhi iris database , and a very high accuracy rate is achieved .
three sample iris images of the dataset used in this work are shown in figure 1 .
the rest of the paper is organized as follows .
section [ sectionii ] describes the features which are used in this work .
section [ sectioniii ] contains the explanation of the classification scheme .
the results of our experiments and comparisons with other works are in section [ sectioniv ] and the paper is concluded in section [ sectionv ] .
extracting good features and image descriptors is one of the most important steps in many computer vision and object recognition algorithms . as a result , many researchers have focused on designing useful features which can be used for a variety of object recognition and image classification tasks .
a good feature should have some degree of invariance with respect to translation , slight rotation and deformation .
there are many popular features and image descriptors which are being used today , including scale invariant feature transform ( sift ) , histogram of oriented gradient ( hog ) , bag of words ( bow ) @xcite-@xcite , etc .
geometrical features and sparsity - based features are also used for some biometric and medical applications in several works @xcite-@xcite . a new algorithm for feature selection for small datasets
is presented in @xcite .
recently , unsupervised feature learning algorithms have been in the spotlight , where the image is fed directly as the input to the deep neural network and the algorithm itself finds the best set of features from the image . for iris recognition ,
various features have been used by several researchers , including wavelet - based features , pca and lda . in this paper , a combined set of two features
is used : some derived from the scattering transform , and the rest from the textural information of iris patterns .
these features are introduced in the following subsections in more detail .
the scattering operator is a locally translation - invariant descriptor which is proposed by stephane mallat and has achieved state - of - the - art recognition accuracy in several computer vision @xcite and audio classification @xcite problems .
a scattering transform computes local image descriptors with a cascade of three operations : wavelet decompositions , complex modulus and a local averaging .
the scattering coefficients are similar to those of the sift descriptor , but they contain more high - frequency information than sift .
as discussed in @xcite , other image descriptors such as sift and multiscale gabor textons can be obtained by averaging the amplitude of wavelet coefficients , calculated with directional wavelets .
this averaging provide some sort of local translation invariance , but it also reduces the high - frequency information .
scattering transform recovers part of the high - frequency information lost by this averaging with co - occurrence coefficients having the similar invariance as those of the scattering transform . in most object recognition tasks , locally invariant features are preferred , since they provide robust representation of images .
they can be seen as the averaged value of gradient orientation . using this averaging , some local deformation and translation will be tolerated .
however , such process will reduce too much high - frequency information and therefore could greatly decrease discriminating capability .
the scattering features provide richer descriptors for complex structures such as corners and multiscale texture variations .
the scattering operator is designed in a way that it preserves the locally invariance property of sift , but it also recovers the lost high - frequency content of the images .
suppose we have a signal @xmath0 .
the first scattering coefficient is the average of the signal and can be obtained by convolving the signal with an averaging filter @xmath1 as @xmath2 .
the scattering coefficients of the first layer can be obtained by applying wavelet transforms at different scales and orientations , and taking the magnitude and convolving it with a low - pass filter @xmath1 as shown below : @xmath3 and @xmath4 denote different scales and orientations and @xmath5 . note that removing the complex phase of wavelet will make these coefficients insensitive to local translation . now to recover the high - frequency information , which is eliminated from the wavelet coefficients of first layer by averaging , we can convolve @xmath6 by another set of wavelet at scale @xmath7 , taking the absolute value of wavelet and taking the average : @xmath8 is negligible at scales where @xmath9 . therefore the coefficients are calculated only for @xmath10 .
the convolution with @xmath1 at the second layer removes high frequencies and yields second - order coefficients which are locally invariant to translation .
these high - frequency information can be restored again by finer scale wavelet coefficients in the next layers . to obtain the scattering coefficients at the @xmath11-th layer
, we have to perform the following procedure iteratively @xmath11 times : @xmath12 the output of scattering transform of the @xmath11-th layer has a size of @xmath13 where p denotes the number of different orientations .
in other words there are @xmath13 transformed images at the output of the @xmath11-th layer .
the transformed images of the first and second layers of scattering transform for a sample iris image are shown in figures 2 and 3 .
these images are derived by applying bank of filters of 5 different scales and 6 orientations .
scattering vector can be thought of as the cascade of convolution with wavelets , non - linear modulus and averaging operators which makes it very similar to the deep convolutional neural network @xcite . to derive scattering features ,
the scattering - transformed images of all layers up to @xmath14 are taken and the mean and variance of these images are calculated as scattering features which results in a vector @xmath15 of size @xmath16 . for mainstream applications , using two or three levels of scattering transform will be enough .
textural , spectral and contextual features are the three fundamental pattern elements in recognition . in human interpretation of color photographs ,
textural features contain the spatial information of intensity variation in a single band @xcite . to mimic the human visual system ,
there are several features introduced to capture textural information of an image . among them , haralick features and local binary pattern ( lbp ) are two major groups of textural features .
haralick textural features are derived from the co - occurrence matrix of image . local binary pattern
are derived based on the relative comparison between each pixel and its neighboring pixels .
there are also various modified versions of lbp features such as transition local binary patterns , direction - coded local binary patterns , volume local binary pattern ( vlbp ) . in this work ,
haralick features are used to capture textural information of the image . to extract haralick features , we first need to derive the co - occurrence matrix .
the co - occurrence matrix measures the distribution of co - occurring intensity values for a given offset . if we represent the image as a two - dimensional function which maps pairs of coordinates to the intensity values , i.e. , @xmath17 , where @xmath18 and @xmath19 , and @xmath20 denotes the set of all possible grayscale levels . then the co - occurrence matrix @xmath21 of image @xmath22 with the offset @xmath23 or @xmath24can be defined as : @xmath25 where @xmath26 denotes the discrete dirac function . it should be noted that the co - occurrence matrix has a size of @xmath27 , where @xmath28 denotes the number of gray levels in the image .
the offset @xmath23 depends on the direction @xmath29 which can be defined as : @xmath30 here we have derived the co - occurrence matrix for the offset @xmath31 . in our work , the textural features are extracted on block level .
each image is divided into non - overlapping blocks of size @xmath32 and their co - occurrence matrices are derived and 14 features are extracted from them .
more details about the derivation of these 14 textural features from co - occurrence matrix is provided in the appendix .
then the features from different blocks are concatenated and formed a longer feature vector .
if an image has a size of @xmath33 , the total number of textural features will be : @xmath34 in our work the textural features are derived in a slightly different way from the original paper @xcite , but they are very similar . here
the co - occurrence matrix is found only for a single pixel horizontal shift ( corresponding to @xmath35 ) .
after derivation of the set of scattering and textural features , we can concatenate them to form the feature vector of each iris image as : @xmath36^\intercal$ ] , where @xmath15 and @xmath37 denote the scattering and textural features respectively . principal component analysis ( pca ) , also known as karhunen - loeve transformation , is a powerful algorithm used for dimensionality reduction @xcite .
given a set of correlated variables , pca transforms them into another domain such the transformed variables are linearly uncorrelated .
this set of linearly uncorrelated variables are called principal components .
pca is usually defined in a way that the first principal component has the largest possible variance and the second one has the second largest variance and so on .
therefore after applying pca , we could only keep a subset of principal components with the largest variance to reduce the dimensionality .
pca can be thought of as fitting a @xmath11-dimensional ellipsoid to a set of data , where each axis of the ellipsoid represents a principal component .
there are also others dimensionality reduction algorithms which are designed based on pca such as kernel - pca and sparse - pca .
pca has many applications in computer vision .
eigenface is one representative application of pca in computer vision , where pca is used for face recognition @xcite . without going into too much detail ,
let us assume we have a dataset of @xmath38 iris images and @xmath39 denote their features .
also let us assume that each feature has dimensionality of @xmath40 . to apply pca , all features need to be centered first by removing their mean : @xmath41 where @xmath42 .
then the covariance matrix of the centered images is calculated : @xmath43 next the eigenvalues @xmath44 and eigenvectors @xmath45 of the covariance matrix @xmath46 are computed .
suppose @xmath44 s are ordered based on their values .
then each @xmath47 can be written as @xmath48 .
we can reduce the dimensionality of the data by projecting them on the first @xmath49 principal vectors as : @xmath50 by keeping @xmath11 principal components , the percentage of retained variance can be found as : @xmath51 . hence one simple way to choose @xmath11 would be to pick a value such that the above ratio is less than @xmath52 , where @xmath52 is usually chosen between 95% to 99% .
there are various classifiers which can be used for this task , including majority voting algorithm , support vector machine , neural network and minimum distance classifier . in this work minimum distance classifier
has been used which is quite popular for template matching problems .
one benefit of minimum distance classifier is that it does not need any training , making it much faster than most of the other classifiers .
as long as the features are discriminative enough to separate different classes , the minimum distance classifier will provide high accuracy , otherwise using other classifiers would be a better option .
minimum distance classifier finds the distance between the features of the training samples and those of an unknown subject , and picks the training sample with the minimum distance to the unknown as the answer . to put it in equation ,
if we show the features of the test subject as @xmath53 and those of the test sample @xmath54 with @xmath55 , the test subject is matched to the sample that satisfies the following : @xmath56\end{gathered}\ ] ] we have used euclidean distance as our distance metric .
this section presents a detailed description of experimental results . before showing the results ,
let us describe the parameter values of our algorithm .
for each image , scattering transform is applied up to two levels with a set of filter banks with 5 scales and 6 orientations , resulting in 391 transformed images . the mean and variance of each scatter - transformed image
are used as features , resulting in 782 scattering features .
the scattering features are derived using the software implementation provided by mallat s group @xcite . to extract textural features ,
each image is divided into 12 smaller blocks and 14 features are derived from any of them which results in a total of 168 textural features . therefore the concatenated feature vector has a length of 950 .
then pca is applied to all features and the first 80 pca features ( retain above 99% of the initial features energy ) are used for recognition .
minimum distance classifier is used for template matching .
we have tested our algorithm on a popular iris database collected by iit delhi .
this database contains 2240 iris images captured from 224 different people .
the images of 21 people ( around 10% ) are used as a validation set to find the optimum value of the parameters of the algorithm , and the rest are used for evaluation . for each person , around half of the images are used for training and the rest for testing .
figure 4 shows the recognition accuracy using different numbers of pca features .
interestingly , even by using few pca features , we are able to get a very high accuracy rate .
as it can be seen , using 80 pca features results in a accuracy rate above 99% , which will not increase by using more pca features .
table 1 provides a comparison of the performance of the proposed scheme and those of other recent algorithms .
the accuracy of the proposed scheme is reported as the highest rate achieved by 80 pca features . as it can be seen , using the combination of scattering and textural features , we are able to outperform previous approaches .
this is mainly due to the richness of both scattering and haralick features which are able to capture high - frequency patterns of irises , providing a very high discriminating power .
one main advantage of this scheme is that , it does not require segmentation of iris from eye images ( although the segmentation could improve the results for some difficult cases ) .
[ h ] haar wavelet @xcite & 96.6% + log gabor filter by kumar @xcite & 97.19% + fusion @xcite & 97.41% + elastic graph matching @xcite & 98% + proposed scheme using 80 pca features & 99.2% + [ tblcomp ] the experiments are performed using matlab 2012 on a laptop with core i5 cpu running at 2.6ghz .
the execution time for the proposed scheme is about 11ms for each image which is fast enough to be used for real - time applications .
this paper proposed a set of scattering and textural features for iris recognition .
the scattering features are extracted globally , while the textural features are extracted locally .
scattering features are locally invariant and carry a great deal of high - frequency information which are lost in other descriptors such as sift and hog .
the high - frequency information provides great discriminating power for iris recognition .
principal component analysis is applied on features to reduce dimensionality .
then minimum distance classifier is used to match new iris images with training images .
this algorithm is tested on a well - known dataset , and a high accuracy rate is achieved which outperforms the previous best results achieved on this dataset . in the future , we will investigate to apply the proposed set of features to more challenging iris datasets and also other biometric recognition problems .
the authors would like to thank stephane mallat s research group at ecole normale superieure for providing the software implementation of scattering transform .
we would also like to thank iit delhi for providing the iris database @xcite .
to find textural features from the co - occurrence matrix , we first need to find the following terms which are used for derivation of haralick features .
@xmath57 denotes the normalized co - occurrence matrix which can be thought as a probability distribution . @xmath58 and @xmath59 denote the marginal probabilities along @xmath60 and @xmath61 . @xmath62 and
@xmath63 denote the probabilities of @xmath64 and @xmath65 .
@xmath66 denotes the entropy of @xmath67 and : @xmath68 using the above terms , the following 14 textural can be derived for each ^2 , \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \text{angular second moment } \\ & f_2= \sum_{k=0}^{n_g-1}k^2 p_{x - y}(k ) , \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \text{contrast } \\ & f_3= \frac{\sum_{i}\sum_{j}ijp(i , j)-\mu_x\mu_y}{\sigma_x\sigma_y } , \ \ \ \ \ \text{correlation } \\ & f_4= \sum_{i}\sum_{j } ( i-\mu)^2p(i , j ) , \ \ \ \ \ \ \ \ \ \ \text{variance } \\ & f_5= \sum_{i}\sum_{j } \frac{1}{1+(i - j)^2 } p(i , j ) , \ \ \ \text{inverse diference moment}\\ & f_6= \sum_{k=2}^{2n_g}k p_{x+y}(k ) , \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \text{sum average } \\ & f_7= \sum_{k=2}^{2n_g}(k - f_6)^2 p_{x+y}(k ) , \ \ \ \ \ \ \ \ \ \ \ \ \text{sum variance } \\ & f_8= -\sum_{k=2}^{2n_g } p_{x+y}(k ) \log(p_{x+y}(k ) ) , \
\text{sum entropy } \\ & f_9= -\sum_{i}\sum_{j } p(i , j ) \log(p(i , j ) ) , \ \
\text{entropy } \ \\ & f_{10}= \sum_{k=0}^{n_g-1}(k-\mu_{x - y})^2 p_{x - y}(k ) , \ \ \ \ \text{diference variance } \\ & f_{11}= -\sum_{k=0}^{n_g-1 } p_{x - y}(k ) \log p_{x - y}(k ) , \ \text{difference entropy } \\ & f_{12}= \frac{hxy - hxy1}{max\{hx , hy\ } } \\~\\ & f_{13}= \sqrt{1-exp{[-2(hxy2-hxy ) ] } } \\~\\ & f_{14}= \sqrt{\text{second largest eigenvalue of } \ q } \\\end{aligned}\ ] ] 1 ak . hrechak and ja .
mchugh , `` automated fingerprint recognition using structural matching '' , pattern recognition 23.8 : 893 - 904 , 1990 . s. minaee , aa . abdolrashidi , `` multispectral palmprint recognition using textural features , '' ieee signal processing in medicine and biology symposium ( spmb ) , 2014 . s. minaee , aa .
abdolrashidi , `` highly accurate multispectral palmprint recognition using statistical and wavelet features , '' ieee signal processing workshop , 2015 .
w. zhao , r. chellappa , pj .
phillips and a. rosenfeld , `` face recognition : a literature survey '' , acm computing surveys ( csur ) 35 , no . 4 : 399 - 458 , 2003 .
kw bowyer , kp hollingsworth and pj flynn , `` a survey of iris biometrics research : 20082010 , '' handbook of iris recognition .
springer london , 15 - 54 , 2013 .
daugman , `` high confidence visual recognition of persons by a test of statistical independence , '' ieee transactions on pattern analysis and machine intelligence , 15.11 : 1148 - 1161 , 1993 .
a. kumar and a. passi , `` comparison and combination of iris matchers for reliable personal authentication , '' pattern recognition , vol .
1016 - 1026 , mar .
farouk , `` iris recognition based on elastic graph matching and gabor wavelets , '' computer vision and image understanding , elsevier , 115.8 : 1239 - 1244 , 2011 . c. belcher and y. du , `` region - based sift approach to iris recognition , '' optics and lasers in engineering , elsevier 47.1 : 139 - 147 , 2009 .
pillai , vm .
patel , r. chellappa and nk .
ratha , `` secure and robust iris recognition using random projections and sparse representations '' , ieee transactions on pami , 33 , no . 9 : 1877 - 1893 , 2011
. z. he , t. tan , z. sun and x. qiu , `` toward accurate and fast iris segmentation for iris biometrics '' , ieee transactions on pami , 31 , no . 9 : 1670 - 1684 , 2009 . s. minaee and y. wang , `` screen content image segmentation using least absolute deviation fitting '' , ieee international conference on image processing , 2015 . s. minaee , h. yu and y. wang , `` a robust regression approach for background / foreground segmentation '' , arxiv preprint arxiv:1412.5126 , 2014 .
m. hosseini , a. fedorova , j. peters and s. shirmohammadi , `` energy - aware adaptations in mobile 3d graphics '' , acm multimedia : 1017 - 1020 , 2012 .
d. lowe , `` distinctive image features from scale - invariant keypoints , '' international journal of computer vision 60.2 , : 91 - 110 , 2004 .
n. dalal and b. triggs , `` histograms of oriented gradients for human detection '' , ieee conference on computer vision and pattern recognition , vol . 1 , 2005 . g. csurka , c. dance , l. fan , j. willamowski and c. bray , `` visual categorization with bags of keypoints '' , eccv workshop , vol . 1 , no . 1 - 22 , 2004 . s. minaee , m. fotouhi and b.h .
khalaj , a geometric approach for fully automatic chromosome segmentation``,signal processing in medicine and biology symposium , ieee , 2014 .
u. srinivas , h. mousavi , c. jeon , v. monga , a. hattel and b. jayarao , ' ' shirc : a simultaneous sparsity model for histopathological image representation and classification , `` isbi , pp .
1118 - 1121 .
ieee , 2013 .
u srinivas , hs mousavi , v monga , a. hattel and b. jayarao , ' ' simultaneous sparsity model for histopathological image representation and classification `` , ieee transactions on medical imaging , 33(5 ) , 1163 - 1179 , 2014 .
s. minaee , y. wang , yw .
lui , ' ' prediction of longterm outcome of neuropsychological tests of mtbi patients using imaging features " , signal processing in medicine and biology symposium ( spmb ) , ieee , 2013 .
j. bruna , s. mallat , `` classification with scattering operators , '' ieee conference on computer vision and pattern recognition , pp.1561 - 1566 , 2011 .
j. andn , s. mallat , `` multiscale scattering for audio classification , '' ismir .
j. bruna , s. mallat , `` invariant scattering convolution networks , '' ieee transactions on pattern analysis and machine intelligence , 35.8 : 1872 - 1886 , 2013 .
haralick , k. shanmugam , and i. dinstein , `` textural features for image classification , '' ieee transactions on systems , man and cybernetics , 6 : 610 - 621 , 1973 .
h. abdi and lj .
williams , `` principal component analysis , '' wiley interdisciplinary reviews : computational statistics 2.4 : 433 - 459 , 2010 .
turk and ap .
pentland , `` face recognition using eigenfaces , '' ieee conference on computer vision and pattern recognition , 1991 . | iris recognition has drawn a lot of attention since the mid - twentieth century . among all biometric features , iris is known to possess a rich set of features .
different features have been used to perform iris recognition in the past . in this paper , two powerful sets of features
are introduced to be used for iris recognition : scattering transform - based features and textural features .
pca is also applied on the extracted features to reduce the dimensionality of the feature vector while preserving most of the information of its initial value .
minimum distance classifier is used to perform template matching for each new test sample .
the proposed scheme is tested on a well - known iris database , and showed promising results with the best accuracy rate of 99.2% . |
in the last decade the great success was obtained in the study of the heavy quarkonium production and decay ( see , for example , refs.@xcite ) .
the investigation of the processes with the heavy quarks is based on the factorization hypothesis @xcite .
the mass of heavy quark @xmath3 is much larger than the scale of strong interactions @xmath4 .
so the presence of the small parameter @xmath5 has allowed one to separate the effects of small and long distances .
the heavy hadron production amplitude may be presented as the product of the partonic part , which can be calculated using the perturbative qcd , and the nonperturbative factor , which describes the free quarks to final hadron transition . in the framework of the nonrelativistic quark model
, this nonperturbative part can be presented through the quarkonium wave function at the origin @xmath6 , which is calculated using the potential method ( see @xcite and references therein ) .
so , the approach based on pqcd and nonrelativistic quark model is known as a nonrelativistic qcd ( nr qcd ) @xcite .
the heavy hadron production via fragmentation prevails in the large transverse momentum region at @xmath7 and hadron colliders @xcite .
recently it was shown @xcite that the fragmentation functions @xmath8 and @xmath9 for quark and gluon to split into a heavy quarkonium with fraction @xmath10 can be calculated within the framework of nr qcd .
the fragmentation functions are the process independent and can be applied to the @xmath7 , photonic and hadronic production of heavy quarkonia . here
we demonstrate that the nr qcd approach can be used for the calculation of the fragmentation function @xmath11 for heavy quark to split into double heavy baryon .
the estimation of the production rates for baryons containing two heavy quarks was done recently on the basis of the quark - hadron duality @xcite as well as on the basis of the pqcd @xcite . in spite of the some differences , it was suggested in refs .
@xcite , that the double heavy baryon production has two step . in the first stage
, there is the heavy quark @xmath12 fragmentation into double heavy diquark @xmath13 in the colour octet state .
the second step consists in the nonperturbative diquark fragmentation into a ( @xmath14 ) baryon .
it was mentioned in ref .
@xcite , that this mechanism of the fragmentation can be factorized into short and long distance contributions as the same in ref .
the normalization of the fragmentation function @xmath11 is determined by the model dependent value of the @xmath13 diquark wave function at the origin @xmath15 as well as by the additional suggestion about nonperturbative diquark fragmentation mechanism into baryon .
note , that the calculation of the @xmath6 for the colour object using potential model is nt grounded and destroys the factorization of the long and short distance effects .
it was also suggested in ref .
@xcite , that the probability of the diquark @xmath13 fragmentation into baryon is equal to unity and the heavy diquark carries all of the momentum of the baryon @xmath16 .
these circumstances raise too high the predicted values of the double heavy baryon production rates .
the model of the quark - hadron duality , which was used in ref.@xcite for the prediction of @xmath0baryon production cross section at @xmath17factory , gives us the opportunity to obtain only upper limit of the production rates and does nt predict baryon spectra .
it was suggested in @xcite that all ( @xmath18)-pairs in the colour octet state with invariant mass from @xmath19 to @xmath20 ( @xmath21 is @xmath22-meson mass and @xmath23 gev ) split to the @xmath18-baryon . however , it is known @xcite that the phenomenological application of the quark - hadron duality method for heavy quarkonium photo- and hadroproduction needs to introduce additional process dependent @xmath24-factor which is equal to @xmath25 for description of the experimental data .
therefore the results of refs . @xcite may be considered as a rough estimation of the double heavy baryon production rates , which need in more detail analysis .
in the present paper we consider a new mechanism of the double heavy baryon production via heavy quark fragmentation based on the hypothesis of the point - like diquarks , which can be produced directly in the hard interactions of quarks and gluons at the short distance @xcite . as it was shown in ref .
@xcite the data on @xmath7 annihilation into hadrons do nt contradict to the existence of the very small diquark @xmath26 containing one heavy and one light quark . in such a way , the heavy quark ( @xmath27 ) can fragment directly into double heavy baryon spin-1/2 or -3/2 catching the scalar or the vector heavy diquark ( @xmath28 ) , correspondingly . in the process of the production quark - diquark system @xmath29 in the colour singlet state
the relevant qcd scale satisfies to the next condition : @xmath30 , where @xmath31 is the heavy diquark mass .
this fact leads to the factorization of the double heavy baryon production amplitude .
the transition of the @xmath29 into final baryon can be described using the nonrelativistic approximation of the potential model , because the system @xmath29 contains two heavy particles .
so , for the calculation of the baryon ( @xmath32 ) wave function at the origin we have used equation @xcite , which effectively takes into account the relativistic effects of the kinematic nature @xcite : @xmath33 where @xmath34 @xmath35[m^2-(m_q - m_d)^2]\over{4m^2}}.\ ] ] the potential of the quark - diquark interaction is taken in the conventional form @xcite : @xmath36 where @xmath37 gev@xmath38 , @xmath39 .
the results of our numerical calculations for the radial part of the wave function at the origin @xmath40 , are presented in tables 1 and 2 .
the fragmentation function @xmath11 at the scale @xmath41 for the production of the baryon @xmath42 , containing the heavy quark @xmath12 and the heavy diquark @xmath26 , is given by the next expression : @xmath43 where @xmath44 is the matrix element for the production of a baryon @xmath42 and antidiquark @xmath45 with the total four - momentum @xmath46 and the invariant mass @xmath47 , @xmath48 is the matrix element for the production of a quark @xmath12 with the same three - momentum @xmath49 .
the lower limit in the integral ( 3 ) is @xmath50 here @xmath51 is the baryon mass , @xmath52 is the baryon transverse momentum in the reference frame where @xmath53 in the limit of @xmath54 : @xmath55 in the axial gauge associated with four - vector @xmath56 : @xmath57 the fragmentation contribution comes only from the feynman diagram for @xmath58 shown in fig.1 .
the gluon couplings to scalar and vector diquarks are presented by the next expressions : @xmath59 @xmath60 f_2(k^2)-\ ] ] @xmath61 where @xmath62 are gell - mann matrices , @xmath63 are the diquark polarization vectors . @xmath64 and @xmath65 are form factors depending on the momentum transfer squared @xmath66 .
the light diquark form factors may be parametrized as follows @xcite : @xmath67 @xmath68 where @xmath69 being the anomalous magnetic moment of the vector diquark , @xmath70 and @xmath71 gev@xmath38 @xcite .
at present time the form factors of the heavy diquarks are unknown and one has to assume a certain dependence on @xmath72 .
the general consideration ( see , for example , ref .
@xcite ) , based on asymptotic qcd picture and the phenomenology , predicts @xmath73 gev@xmath38 , corresponding to a very small heavy diquark .
we try to take these points into account by using the simplest possible expressions for the heavy diquark form factors : @xmath74 @xmath75 @xmath76 where @xmath77 .
it has @xmath78 gev@xmath38 for @xmath79 diquarks and our parametrization is the same as in ref .
note , that such choice of the diquark form factors gives opportunity to obtain the compact analytical expressions for the fragmentation functions . in the case of the heavy quark fragmentation into spin-1/2 baryon it has fusion of the heavy quark @xmath12 and scalar diquark @xmath22 .
after the some obvious simplifications we have obtained : @xmath80 @xmath81\gamma,\ ] ] where @xmath82 , @xmath83 and @xmath84 are the momenta of the quark and the diquark in the baryon , correspondingly , @xmath85 , @xmath86 is the baryon momentum , @xmath87 is the colour factor of the amplitude shown in fig.1 , the dirac spinor @xmath88 is the matrix element for the production of a @xmath12 quark of momentum @xmath89 , @xmath90 is the antidiquark momentum , the spinor @xmath91 describes spin-1/2 baryon in the final state .
the amplitude for heavy quark fragmentation into spin-3/2 baryon , corresponding to fusion of the heavy quark and the vector diquark , can be written as a sum of two parts , which are proportional to @xmath92 and @xmath93 form factors : @xmath94 @xmath95 @xmath96\gamma,\ ] ] @xmath97 @xmath98 ( \hat q+m_q)\gamma.\ ] ] here , the spin - vector @xmath99 describes the spin-3/2 baryon in the final state and satisfies to the following conditions @xcite : @xmath100 @xmath101 in the case of the unpolarized spin-3/2 baryon the summing on the helicity states is carried out by means of the following formula : @xmath102
substituting the amplitudes ( 9 ) and ( 10 ) to the basic formula ( 3 ) we have obtained the fragmentation functions for a heavy quark split into the spin-1/2 and spin-3/2 baryons . omitting the details of the calculation we write here final result for the heavy quark fragmentation function for spin-1/2 baryon : @xmath103 where @xmath104 @xmath105.\ ] ] the fragmentation probability for the production of the spin-1/2 baryon is @xmath106 where @xmath107 @xmath108.\ ] ] the fragmentation function @xmath11 satisfies to the gribov - lipatov - altarelli - parisi ( glap ) evolution equations @xmath122 where @xmath123 the boundary condition on the evolution equation is the initial fragmentation function @xmath124 at the scale @xmath125 .
note , that at leading order in @xmath126 one has : @xmath127 and the evolution equation implies that the fragmentation probability + @xmath128 does not evolve with the scale @xmath129 .
therefore the fragmentation probability is the universal characteristic of the production rates .
the evolution only changes the @xmath10-distribution to smaller values of @xmath10 . the results of calculation for the fragmentation probabilities and the average values of the momentum fraction for @xmath12 quark to split into double heavy spin-1/2 and spin-3/2 baryons are shown in tables 1 and 2 .
the values of @xmath130 are presented at @xmath131 and @xmath132 .
we used the following set of the mass parameters : @xmath133 gev , @xmath134 gev , @xmath135 gev , @xmath136 gev .
the masses of the vector diquarks are greater than corresponding scalar diquarks by 0.1 gev .
the fragmentation functions @xmath137 and @xmath138 , divided by the fragmentation probabilities , are shown in fig . 2 at @xmath131 and @xmath132 .
in the range of accuracy of our model the shape of @xmath139spectra for spin-1/2 and spin-3/2 baryons are the same one to other as well as @xmath139spectra for baryons with strange quark ( @xmath140 ) and without it ( @xmath141 ) .
let compare our results with the estimation @xcite , where it was founded that fragmentation probabilities for heavy quark to split into double heavy baryons are @xmath142 independently on the baryon spin and the flavour content . in our approach
the fragmentation probability depends on baryon spin and flavour .
moreover , there is the relative growth of the production rates of the spin-3/2 baryons for @xmath143-quark fragmentation in comparison with the fragmentation of @xmath144-quark .
we have obtained that @xmath145 and @xmath146 .
our results for the fragmentation probabilities for the spin-3/2 baryons are approximately equal to the results of @xcite , however the fragmentation probabilities for spin-1/2 baryons are smaller about factor 2 . taking into consideration the uncertainties of our calculation , connected with the diquark form factors ( 8)
, we can predict the fragmentation probabilities for the @xmath0 and @xmath1 baryons are about of @xmath147 .
there are more accurate predictions for baryon @xmath10-spectra in the our approach , which are practically independent on the diquark form factors and other parameters .
we have obtained that the average fraction of the baryon momentum @xmath148 for the @xmath0baryons and @xmath149 for the @xmath1baryons at @xmath132 , and @xmath150 and @xmath151 , correspondingly , at @xmath131 .
this results are independent from the baryon spin .
by contrast , in the model , based on @xmath152 fragmentation mechanism corresponding values at @xmath131 are equal : @xmath153 for the @xmath0baryons and @xmath154 for the @xmath1baryons . in conclusion
, we sum the results of our paper . in the framework of nr qcd and a quark - diquark model of baryons we have obtained in the leading order in @xmath126 the fragmentation functions and the probabilities for heavy quark to split into double heavy spin-1/2 and spin-3/2 baryons .
using the qcd evolution equations we have recalculated the fragmentation functions from the initial scale @xmath155 to @xmath132 .
our results can be used for prediction of the double heavy baryon production rates as well as for the description of the energy spectra at @xmath7-collider lep .
contrary to refs .
@xcite we have predicted also few precise effects : the nontrivial spin and flavour dependence of the baryon production rates , the value of the average baryon momentum @xmath130 and it s dependence from the scale @xmath129 as well as from the baryon flavours .
the authors express their gratitude to r.n.faustov , v.o.galkin , v.v.kiselev , and a.k.likhoded for the discussions of the quarkonium physics as well as c.carimalo and d.b.lichtenberg for the information on diquark models .
99 gershtein s.s .
uspekhi phys .
( 1995 ) 3 .
schuler g.a .
preprint cern - th.7170/94 , 1994 .
bodwin g.t .
, braaten e. , lepage g.p .
d51 ( 1995 ) 1125 .
eichten e. , quigg c. preprint fermilab - pub-94/032-t , 1994 .
braaten e. , cheung k. , yuan t.c .
d48 ( 1993 ) r5049 , + braaten e. , yuan t.c .
rev . letters .
71 ( 1993)1673 , + braaten e. , cheung k. phys.rev . d48 ( 1993 ) 4230 .
cheung k. , yuan t.c . physics letters .
b325 ( 1994 ) 481 , + falk a.f . , luke m. , savage m. , wise m. phys.lett .
b312 ( 1993 ) 486 , + saleev v.a .
59 ( 1996 ) 1 , + braaten e. et al .
b333 ( 1994 ) 548 , + kiselev v.v . , likhoded a.k .
, shevlyagin m.v . z.phys .
c63 ( 1994 ) 77 .
kiselev v.v . ,
likhoded a.k .
, shevlyagin m.v .
preprint ihep 94 - 55 , 1994 .
, luke m. , savage m. , wise m. preprint slac - pub-6226 , 1993 .
barger v. , keung w.y .
, phillips r.j.n .
b91 ( 1980 ) 253 .
anselmino m. et al .
v65 ( 1993 ) 1199 .
ekelin s. et al .
d30 ( 1984 ) 2310 .
martynenko a.p .
, saleev v.a .
izv.vuzov.fizika . 5 ( 1995 ) 46 , + martynenko a.p .
vestnik samgu .
1995 , special issue , 109 .
martynenko a.p . ,
faustov r.n .
yad.fiz . 45
( 1987 ) 770 .
anselmino m. , kroll p. , pire b. z.phys .
c36 ( 1987 ) 89 .
kroll p. et al .
b316 ( 1993 ) 546 .
jacob r. , kroll p. phys.lett .
+ carimalo c. , ong s. z.phys .
novozhilov yu.v .
introduction in particle physics , moscow , nauka , 1972 .
( 75.00,47.00 ) ( 20.00,20.00 ) ( 59.00,32.50)(18.00,3.00 ) [ ] ( 58.00,19.00 ) ( 20.00,20.00)(0,0)[cc]@xmath88 ( 72.00,47.00)(0,0)[cc]@xmath86 ( 72.00,12.00)(0,0)[cc]@xmath90 ( 34.00,12.00)(0,0)[cc]@xmath46 ( 39.00,28.00)(0,0)[cc]@xmath158 ( 70.00,28.00)(0,0)[cc]@xmath159 | in the framework of the nonrelativistic qcd and a quark - diquark model of baryons we have obtained the fragmentation functions for heavy quark to split into spin-1/2 and spin-3/2 double heavy baryons . it was predicted the production rates as well as the shape of the energy spectra for the @xmath0 and @xmath1baryons in the region of @xmath2 peak at lep collider .
-2.0 cm -3.0 cm * heavy quark fragmentation into baryons + in a quark - diquark model * + a.p.martynenko , v.a.saleev + samara state university , samara , russia |
observational and theoretical studies have suggested that there are supersonic hydrodynamical or magneto - hydrodynamical turbulent motions in molecular clouds and also in various phases of the ism @xcite .
a number of numerical studies showed , however , that the turbulence in the clouds decays as @xmath1 , with @xmath2 @xcite .
the dissipation time of the turbulence is of the order of the flow crossing time or smaller , even in the presence of strong magnetic fields @xcite .
these numerical experiments suggest that energy input is necessary to maintain the turbulent motion in the molecular clouds .
some energy sources originating in stellar activity have been proposed : stellar winds from young stars @xcite , photoionization @xcite , and supernova explosions @xcite
. however , these processes can not be the main energy sources for keeping the turbulent motion in molecular clouds that do not harbor stellar activities @xcite . in this case
, external shocks caused by distant stellar energy release may produce vortices in the clouds @xcite . in differentially rotating disks with a magnetic field ,
the magneto - rotational instability @xcite is a probable cause for the turbulent motion . @xcite
have studied the mhd turbulence in extended hi disks , applied their model to the extended hi disk of ngc 1058 , and have showed that the observed uniform velocity dispersion may be modeled by the mhd - driven turbulence . in terms of the origin of the non - self - gravitating , pure hydrodynamical turbulence , on the other hand , balbus ,
hawley and stone ( 1996 ) have claimed , on the basis of theoretical and numerical studies , that keplerian disks are linearly and non - linearly stable .
they rejected self - sustained hydrodynamical turbulence with outward angular momentum transport , because the turbulence can not gain energy from the differential rotation .
however , @xcite suggested a steady hydrodynamical turbulence in keplerian disks , which are caused by the pressure - strain tensor .
richard & zahn ( 1999 ) have re - analiezed laboratory experiments of the couette - taylor flow , and concluded that turbulence may be sustained by differential rotation when @xmath3 .
@xcite showed that purely hydrodynamical perturbations can develop initially into either sheared disturbances or coherent vortices ( see also papaloizou & pringle 1985 ) .
however , the perturbations decay and do not evolve into a self - sustained turbulence .
another important mechanism that generates turbulence in a rotating disk is self - gravitational , local or global instability of the gas @xcite .
although there are some simulations of self - gravitating , turbulent gas in a local shearing box ( e.g. v@xmath4zquez - semadeni et al .
1995 ; gammie 2001 ) , the local approximation would not be adequate to study the nature of self - gravity dominated turbulence @xcite .
the global hydrodynamical simulations given by @xcite have revealed spiral unstable modes in the differentially rotating disk , but the development of self - sustained turbulence was not observed . in the this paper , we study the gravity - driven turbulence in galactic disks using two - dimensional , global hydrodynamical simulations .
we solve the basic hydrodynamical equations and the poisson equation numerically , taking into account realistic radiative cooling .
we use the same numerical technique discussed by wada & norman ( 1999,2001 ) ( hereafter wn99 , wn01 ) , in which kpc - scale dynamics of the multi - phase ism can be followed with a sub - pc resolution . wn99 and wn01 suggested that a gravitationally unstable and thermally unstable disk can evolve into a globally quasi - stable disk with quasi - stationary turbulent velocity field . here
we perform simulations with a higher spatial resolution , and make a detailed analysis of the turbulent structure . in 2 , we briefly summarize the numerical method and models , and in 3 , we show that turbulent energy spectra are achieved over a wide dynamic range ( from kpc to pc ) in the direct numerical experiments . in order to study the effects of pure turbulence in the disk
, we here ignore the energy feedback due to supernovae and stellar winds from massive stars @xcite . in 4 , we discuss what is required for realistic models of the ism , and we apply our model to the hi disk of the dwarf galaxy , ngc 2915 . extended hi disks are particularly relevant objects to compare with our models , because they are not significantly affected by the star formation @xcite .
we summarize our results in 5 .
the numerical methods are the same as those described in wn99 and wn01 . here
we summarize them briefly .
we solved the following equations numerically in two - dimensions : @xmath5 & = & \gamma_{\rm uv}- \sigma_g \lambda(t_{\rm g } ) , \label{eq3}\\ \nabla^2 \phi_{\rm sg } & = & 4 \pi g \sigma_g \delta(z ) \label{eqn : poi}\end{aligned}\ ] ] where , @xmath6 are surface density , pressure , and velocity of the gas , the specific total energy @xmath7 $ ] with @xmath8 and a scale height @xmath9 .
the potential has two parameters , @xmath10 ( scale parameter called the ` core radius ' ) and @xmath11 ( the maximum rotation speed ) , and is expressed as @xmath12 where @xmath13 is given as @xmath14 , with @xmath15 km s@xmath16 and @xmath170.2 and 2 kpc .
we also assume a cooling function @xmath18 @xmath19 with solar metallicity and a heating due to photoelectric heating , @xmath20 . a constant scale height ( @xmath21 pc )
is assumed to calculate the cooling rate .
we assume a uniform uv radiation field , equal to the the local galactic uv field @xcite .
however , the uv heating does not significantly affect the dynamics of the ism .
the hydrodynamic part of the basic equations is solved by ausm ( advection upstream splitting method ) , @xcite .
we achieve third - order spatial accuracy with muscl @xcite .
the limiter function is chosen to satisfy the tvd ( total variation diminishing ) condition . in this scheme ,
an explicit numerical viscous term is not required and shocks are captured sharply .
we used @xmath22 equally spaced cartesian grid points covering a 2 kpc @xmath23 2 kpc region around the galactic center .
we also run models using @xmath24 and @xmath25 grid points to evaluate how numerical resolution affects the results .
the poisson equation is solved to calculate the self - gravity of the gas using the fft and the convolution method @xcite .
the second - order leap - frog method is used for the time integration .
we adopt implicit time integration for the cooling term .
the initial condition is an axisymmetric and rotationally supported disk assuming the toomre@xmath0 parameter , @xmath26 .
random density and temperature fluctuations are added at each grid cell in the initial disk .
these fluctuations are less than 1 % of the unperturbed values and have an approximately white noise ( flat power spectrum ) distribution .
as described in wn01 , thermal and gravitational instabilities grow non - linearly over the whole of the disk in a few dynamical time scales .
the whole system shows a quasi - stable state after @xmath27 myr ( 2 rotational periods at @xmath28 kpc ) . as shown in fig .
1 , which is a density field of a part of the disk ( 500 pc @xmath29 500 pc , i.e. 1/16 of the entire simulated region ) , a complicated network of clumps and filaments as well as low density voids are formed .
the density ranges from @xmath30 to @xmath31 @xmath32 pc@xmath33 . the velocity field in the same region in fig .
1 is also very complicated as shown in fig .
2a and 2b , in which @xmath34 and @xmath35 are plotted . comparing to fig . 1 and 2 , we see that regions of converging velocity roughly coincide with dense filaments , and positive and negative vortices are also associated with the filaments .
this shows that local shearing motions are generated along the filaments .
one should note that this turbulent velocity field is maintained in a rotating , self - gravitating gas disk without any explicit energy inputs , such as supernovae or any other artificial driving forces . in order to understand the nature of the turbulent velocity field statistically
, we calculate the energy spectra , @xmath36 from the velocity field , where @xmath37 is the wave number .
we decompose the velocity field into two components , i.e. @xmath38 , where @xmath39 is the compressible velocity field and @xmath40 is the incompressible ( or solenoidal ) field @xcite .
the two components are defined from @xmath41 and @xmath42 , and solenoidal and compressible energy spectra , @xmath43 and @xmath44 are calculated from @xmath40 and @xmath39 , respectively .
figure 3a shows the evolution of the compressible energy spectrum , @xmath44 .
one finds that the compressible part of the velocity field reaches a quasi - steady state in a few rotational periods , where the spectrum shows a double power - law shape with a ` knee ' at @xmath45 ( @xmath46 pc ) .
the kinetic energy is mainly driven at wavenumbers around @xmath47 .
the turbulence inversely cascades to a larger scale ( @xmath48 ) , and simultaneously it also cascades to smaller scales .
the input wave number , @xmath49 , corresponds to about 10 pc , which is the scale of the initial gravitational instability in the cool phase after the initial cooling .
the inverse cascade implies a hierarchical growth of the gravitational instability in the disk .
the energy spectrum of the solenoidal component @xmath43 similarly evolves as that @xmath44 does ( fig .
3b ) , but it has more power especially on the large scale ( @xmath50 or @xmath51 pc ) because of the galactic rotation .
the appearance of the ` hump ' at @xmath52 at @xmath53 myr and the following evolution suggest that eddies are generated on the same scale as the gravitational instability , and they propagate toward large and small scales . the rotational energy on a larger scale cascades downward at @xmath54 .
this process is just like evolution of the gravitational instability in a rotating disk @xcite .
we discuss this below .
a stationary power - law regime in the energy spectrum of the system implies a fully - developed turbulence , and its power - law index represents some statistical aspects of the turbulence . before discussing the power - law indices and the range of the inertial cascade , we first check the simulation for numerical artifacts .
since the turbulence of the ism is a phenomenon over a wide dynamic range , one should be careful about fitting the numerically obtained spectrum with a power - law .
we are using a non - adaptive eulerian grid , therefore the results on smaller scales can be affected by the numerical resolution .
we therefore check the dependence of the turbulent spectrum on the resolution by using @xmath25 and @xmath24 grid points as well as the @xmath22 grid points ( i.e. 0.49 pc resolution ) . in fig . 4 ,
@xmath55 for three cases using @xmath56 and @xmath24 grid points at @xmath57 are plotted .
we find that 1 ) the three runs coincide at @xmath48 within a factor of two , 2 ) the slopes of the ` inertial ranges ' are the same @xmath58 , and 3 ) the power - law indices of the ` dissipation regions ' in the all three models approach to @xmath59 .
the positions of the ` knees ' are shifted by about factor of two , therefore the inertial range is increased by a factor of two if we use twice the number of grid points .
therefore , we can safely conclude that the slope of the inertial range is not strongly affected by the spatial resolution .
in fully developed kolmogorov incompressible turbulence , the steady power - law energy spectrum indicates a steady energy flow from large scales toward smaller scales . in two - dimensional , incompressible , hydrodynamical turbulence , stationary spectra
are found to be proportional to @xmath60 for @xmath61 ( enstrophy , ( i.e. squared vorticity ) cascade regime ) and to @xmath62 for @xmath63 ( energy ` inverse ' cascade regime ) , where @xmath64 is the forcing wavenumber @xcite .
this inverse energy cascade is also observed in numerical simulations of a weakly compressible 2-d turbulence @xcite .
the evolution of the spectrum in fig .
3a look consistent with the kraichnan s picture , with @xmath65 , although our spectrum is shallower ( @xmath66 ) in the inertial range than the kraichnan s prediction . in fig . 5 ,
the final energy spectra ( @xmath67 myr ) , @xmath36 , @xmath43 , and @xmath44 for the model using @xmath22 grid points are plotted .
@xmath44 has a power - law part @xmath68 between @xmath69 and 200 , and @xmath70 between @xmath71 and 200 .
it is known that for shock - dominated , fully - developed turbulence , @xmath72 , and for a weakly shocked turbulence , @xmath73 are expected @xcite .
the present spectrum is shallower than these values , and it implies that the structure of velocity field is not determined only by shocks .
however , it is not clear whether this difference originates from the complexity of the present system , i.e. a non - uniform , rotating , self - gravitating , and radiative turbulence .
even so , it is interesting that the resultant spectra in this complicated system show stationary power laws over inertial ranges as seen in the much simpler systems - uniform , non - selfgravitating , and non - radiative turbulence .
note that the slope will change if the simulations are generalized to include energy feedback from supernovae , which cause strong shocks in the ism ( wn01 ) .
next , we show energy spectra for two models with different rotation curves , i.e. rigid and differential rotation curves ( @xmath74 and @xmath75 kpc ) in fig .
the plot shows that the spectra of the rigid rotation model on large scales ( @xmath76 ) are smaller than those of the differential rotation model by a factor @xmath77 .
on the other hand , the difference of the spectra between the two models are a factor @xmath78 on small scales ( @xmath79 ) .
interestingly , the spectra of the inertial range in the rigid rotation model are steeper ( @xmath80 ) .
it is reasonable that energy extracted from the galactic rotation is more effective when there is a shear field .
we discuss this more quantitatively below .
the above numerical results suggest that the ism in galaxies is dynamic , but in a statistical sense , it is in a ` quasi - steady state ' , whose velocity field resembles fully developed turbulence over a wide dynamic range ( at least three orders of magnitude ) .
the expected velocity dispersions are @xmath81 km s@xmath16 and 0.1 km s@xmath16 for 100 pc and 1 pc scales , respectively .
there are two obvious energy sources that can maintain the turbulence in this system : the shear - driven by the galactic rotation and local tidal field due to the self - gravity of the gas .
besides these , local pressure gradients can also contribute to generate the turbulence .
this is because that the turbulent ism is no longer in pressure equilibrium ( wn01 , gazol et al .
2001 ) . as shown by the energy spectra ( fig .
5 ) , the rotational energy is dominated by that on a large scale ( @xmath82 pc ) . on the other hand ,
turbulence is first driven by the self - gravity of the gas on a scale of 1020 pc as seen in fig .
3 . these energy inputs are in equilibrium with the energy transport due to turbulent decay and losses due to the radiative cooling . in order to see the effect of the gravitational instability on the generation of the turbulence
, we calculate the _ effective _ toomre stability parameter @xmath83 , where the effective velocity dispersion , @xmath84 , is defined by @xmath85 with the velocity dispersion @xmath86 and the sound velocity @xmath87 , the average surface gas density @xmath88 and , the epicyclic frequency , @xmath89 .
all quantities are averaged in local subregions ( 1 kpc@xmath90 and @xmath91 ) .
7 shows the volume - weighted distribution functions of the @xmath92 for three different snapshots .
it is notable that @xmath0 is distributed over a wide range , i.e. four - orders of magnitudes . at @xmath93 myr ,
a large fraction of the volume is in the unstable ( @xmath94 ) state .
note that , although the @xmath95 implies that the axisymmetric mode is linearly unstable , the disk can be unstable to non - axisymmetric modes even with @xmath96 .
the peaks shift to larger values for later times , which means the disk is stabilized . at @xmath97 myr ,
the median is @xmath98 , and the regions where @xmath99 and @xmath100 coexist .
this situation is clearly seen in fig .
8 , which shows the 2-d distribution of @xmath92 and @xmath101 at @xmath97 myr .
the distribution of the stable ( yellow - white ) and unstable regions ( dark red - blue - black ) are very patchy .
it is obvious that the unstable regions correspond to high density regions , which are seen in the density map colored red - yellow .
the disk is more unstable in the outer region ( @xmath102 kpc ) than in the inner region .
the density map shows some global spiral structures , which can be also seen in the @xmath0 map .
the inhomogeneous distribution of @xmath0 appears in the initial a few rotations , and then they evolve into larger structure .
this evolution represents the development of the inhomogeneous structure in density , and the evolution of @xmath36 as shown in fig .
3 is consistent with this result .
the most unstable wave length @xmath103 in the differentially rotating disk is @xmath104330 pc .
this is consistent with the typical size of the unstable regions seen in fig .
8 , which is less than about 50 pc .
the sizes of the unstable regions are larger than those of the dissipational region seen in the energy spectra ( fig .
5 ) , therefore the patchy structure is not strongly affected by the numerical resolution .
the ism in the @xmath94 regions are expected to be collapsed , but this does not mean that all gas components in each unstable region eventually turn into _ one _ high density clump .
in fact , there are substructures in the unstable regions , therefore due to the local angular momentum , the pressure gradient , and the coriolis force , the collapse can not be monotonic . as a result
, the gravitational contraction causes the local velocity dispersion and shock heating , then the regions are stabilized . on the other hand , the stable regions can become unstable due to mass inflow and radiative cooling .
therefore the local structure of the stable and unstable regions are time - dependent , but the entire patchy structure seen in fig . 8 does not change , which is consistent with fact that the energy spectra do not evolve significantly once the system reach the quasi - steady state .
the above results imply that the origin of the turbulent motion in the multi - phase , differentially rotating disk is gravitational instability .
what we observe here is just like the ` local ' gravitational instabilities in a rotating disk as discussed by goldreich & lynden - bell ( 1965a , b ) , although the gas disks here have much more complicated multi - phase structures .
the turbulence is driven by the gravitational instabilities , and probably it is also supplied by energy extracted from the galactic rotation .
this is the same mechanism that drives the mhd turbulence .
as discussed by @xcite , the energy supply rate is given by @xmath105 , where the stress tensor is expressed using alfven velocity @xmath106 and gravitational velocity @xmath107 as @xmath108 ( see also lynden - bell & kalnajis 1972 ) . therefore , even if there is no magnetic field , the gravitational stress tensor works to extract energy from the galactic rotation .
the energy supply rate due to the gravitational instability per unit mass can be estimated as @xmath109 ( where @xmath110 is a scale length of the turbulence ) and @xmath9 is the scale height of the disk . the time scale of the energy supply , i.e. @xmath111 , where @xmath112 is the rotational period , is long enough compared to the galactic life time . in other words , the turbulence can be sustained using a small fraction of the galactic potential energy . the dependence of the energy supply rate on the wavelength can be derived roughly as @xmath113 for the kepler rotation ( @xmath114 ) or @xmath115 for the rigid rotation . since the turbulent energy dissipation rate per unit mass should have the same dependence on the turbulent scale as the energy supply rate does , @xmath116 , or @xmath117 . as a result , we expect that @xmath118 for the kepler rotation , or @xmath72 for the rigid rotation . that is , the power spectrum is steeper in the rigid rotation case than in the differentially rotating case as we see in fig . 6 .
one should note that the turbulence extracts energy from the galactic rotation through self - gravity , and the galactic global shear is not a necessary condition to sustain the turbulence .
in addition to the local instability in self - gravitating disks , interactions with the non - axisymmetric disturbance , such as spiral density waves , can be a mechanism to redistribute energy from the background shear to the local disturbance through the @xmath119 term in the stress tensor @xcite .
this has an analogy with the swing amplification of spiral density waves in differentially rotating , stellar disks @xcite . as seen in fig .
8 , global spiral density waves evolve in the inhomogeneous disks after a few rotational times .
modeling the ism on a galactic scale is a challenging problem in numerical astrophysics .
it is required that realistic simulations include many elementary processes , such as various radiative cooling / heating processes , thermal conduction , interaction between the magnetic field and the ism , self - gravity of the gas , energy feedback from supernovae , etc . moreover
, the simulations should be ultimately three - dimensional and global , that is the whole galactic disk should be solved . in this sense ,
our model and any other past numerical simulations of the ism are still highly idealized ( see a review by v@xmath4zquez - semadeni 2002 ) .
our model is global , and realistic radiative cooling and self - gravity of the gas are taken into account , but the magnetic field , which is important for the evolution of the interstellar turbulence ( see references in 1 ) , is ignored .
therefore we should be careful to apply the present results to the real ism .
on the other hand , most mhd simulations of the ism adopt the local shearing box approximation .
there are many local 2-d models , e.g. self - gravitating ism with radiative cooling ( vazquez - semadeni et al . 1996 ) , and self - gravitating ism with isothermal or adiabatic equation of state @xcite .
some models in 3-d consider radiative cooling and supernova feedback @xcite , but most of them are local , non - selfgravitating , and an isothermal equation of state ( eos ) is assumed @xcite .
there was some global , 3-d mhd simulations with polytropic eos calculated for the galactic central region , but they ignored self - gravity of the gas ( e.g. machida , hayashi , & matsumoto 2000 ) .
using self - gravitating , global hydrodynamical simulations with radiative cooling , on the other hand , wada ( 2001 ) showed that the ism in the galactic central region has the complicated filamentary , multi - phase structure as seen in the 2-d simulations presented here .
the results in the present paper and the past attempts are complementary , and they should be improved with more consistent numerical modeling of the ism in the future .
finally we would like to comment on plausibility of the 2-d approximation for the ism .
the real ism in a galactic disk should behave as three - dimensional turbulence below the scale height of the cold gas ( @xmath120 pc ) .
since the turbulence is driven by local gravitational instability , we expect that the ism in 3-d behaves like 2-d turbulence on larger scales , therefore the inertial range of @xmath121 would also appear for @xmath122 in 3-d ( @xmath123 in 2-d ) . for @xmath124 ,
the energy spectrum would be kolmogorov - like , i.e. @xmath125 , or if it is shock - dominated , @xmath72 would be expected . in a real ism on a smaller scale , the main energy sources would be magneto - hydrodynamical instability and/or energy feedback from stars as well as the energy decay from larger scales .
transition from 2-d to 3-d turbulence and its effect on the energy spectra are now important problems that can be addressed in studies of the gas dynamics in galactic disks , this would be clarified utilizing three - dimensional , global simulations of the ism with magnetic field on a galactic scale .
one might wonder if the spatial resolution in our simulation is meaninglessly fine ( i.e. @xmath126 pc ) , if we apply the model to the real ism whose scale height is @xmath120 pc .
however , as shown in fig .
4 , the transition from the inertia range to the ` dissipative ' part is affected by the numerical resolution . the maximum wave length of the inertia range is about 10 times larger than the grid size . therefore even for 2-d simulations
, the spatial resolution should be much finer than the assumed scale heigh of the disk .
the hi disk in the dwarf galaxy , ngc 2915 , extends to over five times the holmberg radius , and there is no active star formation observed outside the central region .
therefore , this galaxy is relevant to our study of the dynamics of the turbulent gas disk without the influence of the stellar energy feedback .
meurer , mackie , & carignan ( 1994 ) and meurer et al .
( 1996 ) observed ngc 2915 ( @xmath127 mpc , @xmath128 mag , @xmath129 kpc ) using the atca ( australia telescope compact array ) with a linear resolution is 640 pc and the aat ( anglo - astrallia telescope ) .
the total mass of the hi disk is @xmath130 and @xmath131 kpc . the integrated star - formation rate is 0.05 @xmath132 yr@xmath16 , and most of the star - formation is near the center of the galaxy .
the hi intensity map shows that there are spiral arms extending well beyond the optical extent .
the isovelocity contours show that the disk is rotating with a small amount of non - circular motion .
the line - of - sight velocity dispersion of the hi is @xmath133 km s@xmath16 inside the optical radius , and it is @xmath134 km s@xmath16 in the extended hi disk . the central question is what causes this large velocity dispersion in the extended hi disk ?
we apply the same numerical method in 2 to model the hi disk in ngc 2915 to obtain the radial distribution of the velocity dispersion in a non - star forming disk .
we scale the potential model ( eq.(5 ) ) and the gas disk to represent the hi disk of ngc 2915 .
two rotation curves are assumed for an axisymmetric component of the external potential : 1 ) the potential derived from the hi rotation curve ( model a in fig .
16 of meurer et al . 1996 ) , where the core radius @xmath10 in eq.(5 ) is 4 kpc , and 2 ) models with a smaller core radius ( @xmath74 kpc ) .
the latter model was chosen , because the hi rotation curves derived from tracing peak intensities of position - velocity maps do not necessarily represent _ true _ mass distribution of the galaxies , especially for the central part @xcite .
sofue et al .
( 1999 ) claimed that the true central rotation curves tend to be steeper than those implied from the peak intensities in pv maps .
ngc 2915 has a central optical bar .
the observed spiral patterns could be resonance - driven structures by this central bar .
the velocity dispersion in the disk is generally enhanced by the non - axisymmetric potential and its resonances .
the pattern speed of the bar can be directly measured @xcite , and it is @xmath135 km s@xmath16 kpc@xmath16 @xcite . in order to investigate the effects of the central bar , we also performed runs with a non - axisymmetric potential ( bar - potential ) with the same pattern speed as observed ( bureau et al .
1999 ) , and where the length of the bar is taken to be equal to the core radius @xmath10 .
the non - axisymmetric part of the potential is assumed to be in the form @xmath136 where @xmath137 is given as @xmath138 the parameter @xmath139 represents the strength of the bar @xcite .
the gas is initially distributed in an axisymmetric , 15 kpc radius disk with a exponential - like surface density profile to resemble the observed hi distribution ( fig . 15 in meurer
et al . 1996 ) .
we have run models with parameters as @xmath140 km s@xmath16 kpc@xmath16 , @xmath141 , and @xmath74 and 4 kpc .
the velocity dispersions of an axisymmetric model @xmath142 ( @xmath143 kpc ) at @xmath144 gyr are plotted as a function of the radius in fig .
the velocity field is sampled at every 20 grid points ( i.e. 292 pc ) for @xmath145 or @xmath146-directions , and they are averaged in the same size as the observed beam size to calculate the dispersion .
we find that the dispersion is about a factor 4 smaller than the observed value ( 20 - 40 km s@xmath16 ) at @xmath147 kpc , a factor 2 smaller than the observed 8 km s@xmath16 at @xmath148 kpc .
the velocity dispersions of a bar model @xmath149 ( @xmath150 kpc , @xmath151 km s@xmath16 kpc@xmath16 ) is plotted in fig .
the central velocity dispersion in this model is @xmath152 km s@xmath16 , and the it is about 34 km s@xmath16 in the outer disk .
the central velocity dispersion is comparable with the observed value , but the dispersion in the outer disk is still factor 2 smaller than observations .
10 shows @xmath36 for the two models @xmath142 and @xmath149 .
both spectra show double power - law shapes , @xmath118 and @xmath153 in @xmath154 for models @xmath142 and @xmath149 , respectively .
the slope of the axisymmetric model is comparable to that in fig .
5 . the steeper slope in the bar model would be due to stronger shocks caused by the bar .
the spectra also shows that the bar model has a factor of 2 - 10 larger energy than the axisymmetric model in the most scales . in fig .
11 and 12 , we plot the density and pv maps of the bar model @xmath149 .
two major spiral arms , which are formed from many spirals and clumps , are evident as well as an inhomogeneous compact disk .
the pv diagram shows a large velocity dispersion at @xmath155 kpc , which corresponds to the clumpy nuclear disk .
the toomre s @xmath0 parameter estimated by the observations is 59 @xcite .
however , this does _ not _ mean that there are no local gravitationally unstable regions in the hi disk .
the beam size in the observations is much larger than the typical size of the unstable regions .
therefore , we can observe the global stability of the hi disk , but probably the local instability is observed as the velocity dispersion in the beam size .
after exploring the parameter space ( @xmath156 , and @xmath10 ) , we conclude that a set of parameters , i.e. @xmath157 km s@xmath16 kpc@xmath16 , and @xmath158 kpc , reasonably reproduces the observed radial distribution of the velocity dispersion , and the hi morphology
. however , the observed value of the velocity dispersion in the extended disk is about factor two larger than that in the best - fit model @xmath149 .
this might be due to the limitation of our model ( 4.1 ) , or because of the following reasons .
10 suggests that stronger bars can enhance the velocity dispersion in the disk by about factor of two .
if the observed velocity dispersion is a result of _ hidden _ gravitational instability in the hi disk , the dispersion could be larger in a more massive disk .
that is , the observed hi mass does not necessarily represent the total mass of the gas in the galaxy .
there might be a component of high density , compact molecular clouds in the extended hi disk , which could be observed by the next - generation radio interferometer , alma .
dynamics and structure of the interstellar medium ( ism ) in galactic disks are numerically studied , using high - resolution , 2-d hydrodynamical simulations with a large dynamic range . in 3 , we showed that the velocity fields of the ism in a steady - state show turbulent - like spectral energy distributions which are natural consequences of the non - linear development of gravitational instabilities of the ism .
the energy spectra are obtained over three - orders of magnitude in the wave number .
the model using @xmath22 grid points shows that the spectrum for the compressible and incompressible ( solenoidal ) part of the velocity field in the inertia ranges are fitted approximately as @xmath159 and @xmath160 , respectively .
we found that the turbulent energy spectra are achieved in a few dynamical time - scales , and they subsequently remain in a quasi - steady state .
we do not include energy feedback from supernovae or other external driving forces to produce and maintain turbulence .
the turbulence is self - maintained over the wide dynamic range . from analyzing the temporal evolution of the energy spectra and the density structure
, we suspect that energy input from self - gravitational interaction of the ism and the galactic rotation drives the turbulence balancing the turbulent energy decay and the radiative cooling . in 4
, we compared our global ism models with hi observations of ngc 2915 . a large velocity dispersion ( @xmath134 km s@xmath16 ) is observed in the extended hi disk of ngc 2915 , even though there is no active star formation .
we found that the observed velocity dispersion and its radial distribution in ngc 2915 can be quantitatively understood as the gravity - driven turbulence , but the galactic potential requires an additional non - axisymmetric potential with a rotation speed of @xmath161 km s@xmath16 kpc@xmath16 or the total gas mass should be larger than the observed hi mass .
the authors thank fabian heitsch for his valuable comments on the draft .
numerical computations were carried out on fujitsu vpp5000 at naoj .
kw is supported in part by grant - in - aids for scientific research ( no .
12740128 ) of jsps .
balbus , s. a. , hawley , j. f. , & stone , j. m. 1996 , , 467 , 76 balbus , s. a. & hawley , j. f. 1991 , , 376 , 214 balbus , s. a. & papaloizou , j. c. b. 1999 , , 521 , 650 bureau , m. , freeman , k. c. , pfitzner , d. w. , & meurer , g. r. 1999 , , 118 , 2158 dahlburg , j. p. , dahlburg , r. b. , gardner , j. h. , & picone , j. m. 1990 , phys .
fluid , 2 , 1481 dickey , j. m. , hanson , m. m. , & helou , g. 1990 , , 352 , 522 franco , j. & carramiana , a. 1999 , interstellar turbulence , cambridge univ .
press , cambridge gazol , a. , v ' azquez - semadeni , e. , s ' anchez - salcedo , f. j. , & scalo , j. 2001 , , 557 , l121 gammie , c. f. 2001 , , 553 , 174 gerritsen , j. p. e. , icke , v. , 1997 , a&ap , 325 , 972 godon , p. & livio , m. 1999 , , 521 , 319 goldreich , p. & lynden - bell , d. 1965a , , 130 , 97 goldreich , p. & lynden - bell , d. 1965b , , 130 , 125 goldreich , p. & sridhar , s. 1995 , , 438 , 763 goldreich , p. & sridhar , s. 1997 , , 485 , 680 hockney , r. w. , eastwood , j. w. 1981 , computer simulation using particles ( new york : mcgraw hill ) julian , w. h. & toomre , a. 1966 , , 146 , 810 kato , s. & yoshizawa , a. 1997 , , 49 , 213 kim , w. & ostriker , e. c. 2001 , , 559 , 70 .
kornreich , p. & scalo , j. 2000 , , 531 , 366 korpi , m. j. , brandenburg , a. , shukurov , a. , & tumominen , i. 1999 , , 350 , 230 kraichnan , r. h. 1967 , phys .
fluids , 10 , 1417 larson , r. b. 1981 , , 194 , 809 laughlin , g. , korchagin , v. , adams , f.c . , 1998 , apj , 504 , 945 liou , m. , steffen , c. , 1993 , j.comp.phys . , 107,23 lynden - bell , d. & kalnajs , a. j. 1972 , , 157 , 1 mac low , m. , klessen , r. s. , burkert , a. , & smith , m. d. 1998 , physical review letters , 80 , 2754 mac low , m. , apj 524 , 169 machida , m. , hayashi , m. r. , & matsumoto , r. 2000 , , 532 , l67 .
mccray , r. & snow , t. p. 1979
, , 17 , 213 mckee , c. f. 1989 , , 345 , 782 mckee , c. f. & ostriker , j.p .
1977 , apj,218,148 meurer , g. r. , mackie , g. , & carignan , c. 1994 , , 107 , 2021 meurer , g. r. , carignan , c. , beaulieu , s. f. , & freeman , k. c. 1996 , , 111 , 1551 norman , c. a. , & ferrara , a. , 1996 , , 467 , 280 norman , c. a. , & silk , j. 1980 , , 238 , 158 ostriker , e. c. , stone , j. m. , & gammie , c. f. 2001 , , 546 , 980 passot , t. , pouquet , a. , & woodward , p. 1988
, , 197 , 228 papaloizou , j. c. b. & pringle , j. e. 1985 , , 213 , 799 papaloizou , j. c. b. & savonije , g. j. 1991 , , 248 , 353 quiroga , r. j. 1983 , , 93 , 37 richard , d. & zahn , j. 1999 , , 347 , 734 .
sellwood , j. a. & carlberg , r. g. 1984 , , 282 , 61 sellwood , j. a. & balbus , s. a. 1999 , , 511 , 660 sofue , y. , tutui , y. , honma , m. , tomita , a. , takamiya , t. , koda , j. , & takeda , y. 1999 , , 523 , 136 stone , j. m. , ostriker , e. c. , & gammie , c. f. 1998 , , 508 , l99 tremaine , s. & weinberg , m. d. 1984 , , 282 , l5 van leer , b. , 1977 , j.comp.phys . , 32 , 101 v@xmath4zquez - semadeni , e. , pasot , t. & pouquet , a. 1995 , apj , 441 , 702 v@xmath4zquez - semadeni , e. , pasot , t. & pouquet , a. 1996 , apj , 473 , 881 v@xmath4zquez - semadeni , e. 2002 , in `` seeing through the dust : the detection of hi and the exploration of the ism in galaxies '' , eds .
r.taylor , t. landecker , & a. willis , asp , san francisco , in press ( astro - ph/0201072 ) wada , k. 2001 , , 559 , l41 wada , k. & koda , j. 2001 , pasj , 53 , 1163 wada , k. & norman , c. a. 1999 , , 516 , l13 ( wn99 ) wada , k. & norman , c. a. 2001 , , 547 , 172 ( wn01 ) williams , j. p. & blitz , l. 1998 , , 494 , 657 | high - resolution , 2-d hydrodynamical simulations with a large dynamic range are performed to study the turbulent nature of the interstellar medium ( ism ) in galactic disks .
the simulations are global , where the self - gravity of the ism , realistic radiative cooling , and galactic rotation are taken into account . in the analysis undertaken here , feedback processes from stellar energy source are omitted .
we find that the velocity field of the disk in a non - linear phase shows a steady power - law energy spectrum over three - orders of magnitude in wave number .
this implies that the random velocity field can be modeled as fully - developed , stationary turbulence .
gravitational and thermal instabilities under the influence of galactic rotation contribute to form the turbulent velocity field .
the toomre effective @xmath0 value , in the non - linear phase , ranges over a wide range , and gravitationally stable and unstable regions are distributed patchily in the disk .
these results suggest that large - scale galactic rotation coupled with the self - gravity of the gas can be the ultimate energy sources that maintain the turbulence in the local ism .
we find that our models of turbulent rotating disks are consistent with the velocity dispersion of an extended hi disk in the dwarf galaxy , ngc 2915 , where there is no prominent active star formation .
numerical simulations show that the stellar bar in ngc 2915 enhances the velocity dispersion , and it also drives spiral arms as observed in the hi disk . |
rs cvn - type stars are binary systems where the most massive primary component is a g - k giant or subgiant and the secondary is a subgiant or dwarf of spectral classes g to m. these systems are well known due to their strong chromospheric plages , coronal x - ray emission , and strong flares in the optical , uv , radio , and x - ray .
most long - term stellar activity studies of rs cvn stars are derived from the easily detected optical photometric variations produced by their long - lived large spots . in most cases ,
the mean magnitude reflects a stellar activity cycle similar to the 11-year solar one . on the other hand
, the peak - to - peak magnitude shows a shorter cycle , called flip - flop cycle , which reflects the non axisymmetrical redistribution of the spotted area on the stellar surface .
the iue database provides a large number of uv high and low resolution spectra of these type of stars . furthermore , the iue satellite monitored these stars continuously during several seasons . in the present work
, we have measured the mg ii line - core fluxes on the iue low and high resolution spectra of three rs cvn - type stars ( ii peg , ux ari and v711 tau ) and then converted the fluxes to the mount wilson index @xmath1 . for each star
, we analyzed the mean annual index @xmath2 with the lomb - scargle periodogram @xcite to search for a long - term chromospheric cycle .
following daz et al . , we obtained the rotational modulation of the @xmath1-index for several seasons by fitting the light - curves with a harmonic function .
we analyzed the amplitude of each curve with the lomb - scargle periodogram to search for a chromospheric flip - flop cycle .
ii peg ( hd 224085 ) is a single - lined rs cvn - binary system composed by a k2iv star and an unseen companion of an estimated spectral class m0-m3v . in fig .
[ fig.hd224085 ] we plot the mount wilson index @xmath1 derived from the iue high and low resolution spectra obtained between 1979 and 1995 .
derived from the iue high resolution spectra ( @xmath3 ) , and from the iue low resolution spectra ( @xmath4 ) .
arrows indicate the seasons for which we analyzed the modulation .
, scaledwidth=65.0% ] to search for cyclic patterns in the chromospheric data , we first studied the mean annual index @xmath5 of the data plotted in fig .
[ fig.hd224085 ] as a function of time with the lomb - scargle periodogram .
we obtained a peak at 7741 days ( @xmath021 years ) with a false alarm probability ( fap ) of 35% .
secondly , we analyzed the rotational modulation of the index @xmath1 during the seasons indicated with arrows in fig .
[ fig.hd224085 ] . to do so
, we phased each season s light - curve @xmath1 vs. time with the 6.724-day rotation period ( see fig .
[ fig.hd224085_curvas ] ) and we fitted each set of data with the harmonic function : @xmath6 .
we analyzed the amplitude @xmath7 against time with the lomb - scargle periodogram and we obtained a cyclic pattern of 3310 @xmath8 253 days ( 9.07 @xmath8 0.69 years ) with a fap of 31% .
this periodic behaviour seems to be well correlated with the flip - flop cycle with a period of 9.30 years obtained by berdyugina and tuominen for the spot activity .
ux ari ( hd 21242 ) is a rs cvn - type system composed by a k0iv star and a g5v companion in a 6.483-day orbit . in fig .
[ fig.hd21242 ] we plot the index @xmath1 obtained for this star between 1978 and 1996 .
, scaledwidth=65.0% ] we analyzed the mean annual @xmath2 of the indexes plotted in fig .
[ fig.hd21242 ] as a function of time with the lomb - scargle periodogram .
we obtained a period of 2180 @xmath8 32 days ( @xmath06 years ) with a fap of 29% .
v711 tau ( hr 1099 , hd 22468 ) is one of the most active rs cvn non - eclipsing spectroscopic binary system , consisting of a k1 subgiant primary and a g5 dwarf secondary in a 2.837 day - orbit @xcite . in fig .
[ fig.hd22468 ] we plot the index @xmath1 for this star between 1978 and 1995 .
we analyzed the mean annual @xmath2 as a function of time with the lomb - scargle periodogram and we obtained a peak at 6589@xmath81170 days with a fap of 11% .
this period of @xmath9 years is in agreement with the ones reported in the literature . on the other hand
, we analyzed the amplitude of the rotational modulation of the datasets indicated with arrows in fig .
[ fig.hd22468 ] and we obtained a cyclic pattern in the amplitude @xmath10 of the curves plotted in fig .
[ fig.hd22468_curvas ] of period 1207@xmath8 45 days ( @xmath0 3.3 years ) of 23% fap , which is consistent with the one obtained by lanza et al . for the spot activity . | we study the mount wilson indices we obtained indirectly from iue high and low resolution spectra of the rs cvn - type systems ii peg ( k2iv ) , ux ari ( k0iv+g5v ) and v711 tau ( k1iv+g5v ) , extensively observed by iue from 1978 to 1996 .
we analyze the activity signatures , which correspond to the primary star , with the lomb - scargle periodogram . from the analysis of v711 tau data
, we found a possible chromospheric cycle with a period of 18 years and a shorter @xmath03 year cycle , which could be associated to a chromospheric flip - flop cycle .
the data of ii peg also suggest a chromospheric cycle of @xmath021 years and a flip - flop cycle of @xmath09 years .
finally , we obtained a possible chromospheric cycle of @xmath06 years for ux ari .
address = instituto de astronoma y fsica del espacio , conicet - uba , buenos aires , argentina . |
ising models of ferromagnetism , with `` spins '' @xmath0 that take on two values @xmath1 , provide simple examples of systems which undergo a continuous phase transition from a high temperature disordered state to a low temperature state which spontaneously breaks the global symmetry @xmath2 .
the vicinity of this continuous transition poses a challenge to monte carlo methods that rely on local updates . in this critical regime ,
the spin correlation length becomes very large , and local updates are unable to significantly change the state of the system . in such ferromagnetic ising models , this `` critical slowing - down '' of local updates
can be combated using the well - known wolff or swendsen - wang cluster algorithms.@xcite when the interactions between the spins are antiferromagnetic and the underlying lattice non - bipartite , the geometry of the lattice causes these antiferromagnetic interactions to compete with each other .
this `` geometric frustration '' is often associated with a macroscopic degeneracy of minimum exchange - energy configurations .
this can lead to interesting liquid - like phases at intermediate temperatures .
subleading further - neighbour interactions can destabilize this liquid to produce complex patterns of low temperature order .
the standard wolff construction typically fails to yield a satisfactory cluster algorithm in the low temperature regime of interest in such frustrated systems , a result that can be understood in terms of the percolation properties of the wolff clusters.@xcite generalizations@xcite of the wolff cluster construction procedure , which build clusters by defining a percolation process involving larger units of the lattice ( typically , the elementary plaquettes of the lattice ) , have also been explored with some success for the fully frustrated ising model with nearest - neighbour antiferromagnetic exchange on square and triangular lattices .
a ) the triangular lattice is shown in black and its dual honeycomb lattice is shown in blue .
b ) the kagome lattice is shown in black and its dual dice lattice is shown in blue . basis sites and bravais
lattice translation vectors are also marked as appropriate .
, width=317 ] in ferromagnetic models , a `` dual worm '' approach has also been used as an alternative to the standard wolff cluster construction.@xcite this approach uses a worm construction to effect a nonlocal update in the bond - energies along a closed loop . when transformed back to spin variables , this dual worm construction yields a procedure for constructing a spin cluster which can be flipped with probability one . recently
, such a dual worm algorithm has also been formulated for some frustrated systems.@xcite .
these generalizations feature a nonzero rejection rate that is required in order to preserve detailed balance . here ,
we introduce two new cluster algorithms for frustrated triangular lattice ising models with arbitrary two - spin couplings that extend up to next - next - nearest neighbours .
both algorithms use the dual worm framework of ref . .
at the heart of these two algorithms are two new worm construction protocols that both preserve detailed balance _ without any rejection of completed worms both at @xmath3 and @xmath4_. this makes both cluster algorithms very efficient in the entire temperature range of interest in such frustrated magnets . in the limiting case of @xmath5 , or
equivalently , in the @xmath6 limit with @xmath7 and @xmath8 ( with constant @xmath9 and @xmath10 ) , one of these algorithms , which we dub the `` dep '' algorithm ( since it involves deposition , evaporation , and pivoting of dual dimer variables ) , reduces to a previously - used@xcite honeycomb lattice implementation of the well - known worm algorithm for interacting dimer models@xcite . in the same limit , our other algorithm , which we dub the `` myopic '' algorithm since it ignores the environment variables at alternate steps of the worm construction , reduces to the honeycomb lattice implementation of an approach developed previously as an alternative@xcite to the standard procedure@xcite for constructing worm updates for generalized link - current models .
this myopic algorithm generalizes readily to other frustrated systems , including kagome lattice ising antiferromagnets with two - spin couplings again extending up to next - next - nearest neighbours . in the limiting case of @xmath5 on the kagome lattice ( equivalently , in the @xmath6 limit with @xmath7 and @xmath8 with constant @xmath9 and @xmath10 ) , this myopic algorithm reduces to an approach used in previous work to simulate an interacting dimer model on the dice lattice.@xcite in this paper , we present a detailed characterization of both these cluster algorithms in the low temperature power - law ordered intermediate state associated with the two - step thermal melting of three - sublattice order in such triangular lattice ising models .
similar results are also obtained in the kagome lattice case with the myopic algorithm .
we demonstrate that both cluster algorithms have a significantly smaller dynamical exponent @xmath11 compared to that of the standard metropolis algorithm .
interestingly , we also find that the dependence of the dynamical exponent on the equilibrium anomalous dimension @xmath12 is quasi - universal in a sense that we attempt to make precise in this work .
the rest of this paper is organized as follows : in section [ models ] , we introduce the ising models of interest to us and summarize the basic physics of three - sublattice ordering in such triangular lattice antiferomagnets . in section [ algorithms ] , we provide a detailed description of the two cluster algorithms developed here , and present benchmarks establishing the correctness of the procedures used . in section .
[ performance ] , we provide a characterization of the performance of our cluster algorithms , and compare this performance to that of the standard metropolis algorithm .
we conclude in section .
[ outlook ] with a brief discussion .
ising antiferromagnets@xcite on frustrated lattices such as the triangular and the kagome lattice provide paradigmatic examples of the effects of geometric frustration in low dimensional magnets . in such systems ,
the behaviour at low temperature is governed by the interplay between the macroscopic degeneracy of configurations that minimize the nearest - neighbour antiferromagnetic exchange energy , and subleading energetic preferences imposed by weaker further - neighbour interactions .
the classical hamiltonian for these model systems on the triangular and kagome lattices may be written as @xmath13 where @xmath14 , @xmath15 , and @xmath16 denote nearest neighbour , next - nearest neighbour , and next - next - nearest neighbour links of the lattice in question , and @xmath17 are the ising spins at sites @xmath18 of this lattice . in our convention , @xmath19 corresponds to an antiferromagnetic coupling , while @xmath20 corresponds to a ferromagnetic coupling . in the rest of this paper , @xmath21 is assumed positive and equal to @xmath22 . when @xmath23 , the nearest - neighbour model does not order on either lattice even in the zero temperature limit , providing an example of a classical spin liquid state . on the triangular lattice ,
the correlation length grows exponentially with decreasing temperature , reflecting the fact that the @xmath3 spin - liquid , which involves an average over the ensemble of minimum nearest - neighbour exchange energy states , is characterized by power - law spin correlations at the three - sublattice wavevector @xmath24.@xcite on the kagome lattice , @xmath25 with @xmath23 remains a short - range correlated spin liquid all the way down to zero temperature.@xcite when @xmath26 is negative , such magnets tend to develop three - sublattice spin order at low temperature . in this ordered state , the spins freeze in a pattern that is commensurate with the three - sublattice decomposition of the underlying triangular bravais lattice .
this parameter regime has attracted some interest earlier in the context of spatial ordering of monolayer adsorbate films on substrates with triangular symmetry,@xcite and in the context of `` artificial spin - ice '' , _
i.e. _ honeycomb networks of micromagnetic wires which can be modeled in terms of the kagome lattice ising antiferromagnet with further neighbour couplings.@xcite in both these examples , the three - sublattice order is of the _ ferrimagnetic _ type , _
i.e. _ , it is accompanied by a small net moment . as a computationally challenging regime in which to test our algorithms , we focus here on this ferrimagnetc three - sublattice ordered state that is stabilized at low enough temperature by a nonzero ferromagnetic @xmath26 ( @xmath27 ) on both lattices@xcite ( for a simple caricature of this state on both lattices , see fig . 1 of ref . ) . on both lattices
, there is a large range of parameters for which this three - sublattice order melts in a two - step manner on heating,@xcite via an intermediate phase with power - law three - sublattice order corresponding to a temperature dependent power - law exponent @xmath28 . in our work here , we test our algorithms in this extended critical region and extract the values of the dynamical critical exponents that characterize our algorithms .
however , we re - emphasize that the algorithms developed here have much wider applicability , and work well for arbitrary values of @xmath26 and @xmath29 on both lattices when the nearest - neighbour interaction @xmath21 is positive .
the first step of the myopic worm construction on the dual honeycomb lattice : a start site @xmath30 is chosen randomly .
the tail of the worm remains static at this start site until the worm construction is complete . in this first step , the head of the worm moves to one of the three neighbours of the start site with probability @xmath31 , regardless of the local dimer configuration .
the neighbour thus reached is our first vertex site @xmath32 .
viewed from the point of view of this vertex , the start site @xmath30 is the first entry site @xmath33 .
, width=317 ] we begin with a summary of the dual representation of the frustrated ising antiferromagnet on the triangular lattice : one represents every configuration of the triangular lattice ising model in terms of configurations of dimers on links of the dual honeycomb lattice , with a dimer placed on every dual link that intersects a _ frustrated _ nearest neighbour bond of the triangular lattice ( fig . [ lattice ] ) . for our purposes here ,
a frustrated bond of the triangular lattice is one that connects a pair of spins pointing in the same direction . when @xmath34 ( corresponding to the interesting case of frustrated antiferromagnetism ) , this implies that every _ minimally frustrated _ spin configuration , which minimizes the nearest - neighbour exchange energy by ensuring that every triangle of the triangular lattice has exactly one frustrated bond , corresponds to a defect - free dimer cover of the dual honeycomb lattice , in which there is exactly one dimer touching each dual site of the honeycomb lattice . at non - zero temperature
, more general configurations also contribute to the partition sum .
these have a nonzero density of _ defective triangles _ , _
i.e. _ triangles in which all three spins are pointing in the same direction . in dual language , these correspond to honeycomb lattice sites with three dimers touching the site .
thus , in dual language , the configuration space at nonzero temperature is that of a generalized honeycomb lattice dimer model , with either one or three dimers touching each dual site .
this dimer model inherits boundary conditions from the original spin model : we choose to work with @xmath35 samples with periodic boundary conditions on the ising spins along two principal directions @xmath36 and @xmath37 of the triangular lattice .
this translates to global constraints on the dual description which are spelled out in detail when we describe our algorithm .
all of this generalizes readily to the kagome lattice antiferromagnet .
the idea is again to work with the dual representation in terms of a generalized dimer model on the dual lattice .
the dual lattice is now the dice lattice , which is a bipartite lattice with one sublattice of three - coordinated sites and a second sublattice of six - coordinated sites ( fig .
[ lattice ] ) .
every spin configuration on the kagome lattice corresponds to a dimer configuration on the dice lattice , with either one or three dimers touching the three - coordinated sites , and an even number of dimers touching the six - coordinated sites .
as before , a frustrated bond is one that connects a pair of nearest neighbour spins pointing in the same direction , and is represented by a dimer on the dual link that is perpendicular to this bond .
minimally frustrated spin configurations , that minimize the nearest - neighbour exchange energy , now correspond to dimer configurations with exactly one dimer touching each three - coordinated dice lattice site .
periodic boundary conditions of the @xmath35 spin system again translate to global constraints ( see below ) . the probabilistic step of the myopic algorithm : after arriving at the vertex site @xmath38 from the entry site @xmath39 we choose to exit via @xmath40 ( which is one of the neighbours of @xmath38 ) with probability given by the probability table @xmath41 .
a ) when the spin interactions extend upto next - next - nearest neighbours on the kagome lattice , knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath43 and @xmath44 , @xmath45 and @xmath46 suffices to compute entries of the table @xmath41 .
b ) when the spin interactions extend upto next - next - nearest neighbours on the triangular lattice , knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath47 and @xmath44 , @xmath45 and @xmath46 suffices to to compute entries of @xmath41 .
, width=317 ] the _ myopic _ step of the myopic worm algorithm on the dual honeycomb lattice : after arriving at an exit site @xmath40 from a vertex site @xmath38 , the next vertex site @xmath48 is chosen to be one of the two _ other _ neighbours of @xmath40 with probability @xmath49 . viewed from this new vertex site @xmath48 , @xmath40 becomes the entry site @xmath50 .
, width=317 ] the dual worm approach,@xcite on which both cluster algorithms are based , is rather simple to explain in general terms : one first maps the spin configuration of the system to the corresponding dual configuration of dimers .
each dimer configuration is thus assigned a boltzmann weight of the `` parent '' spin configuration from which it was obtained .
next , one updates the dual dimer configuration in a way that preserves detailed balance . in this way , one obtains a new dimer configuration , which is then checked to see if it satisfies certain global winding number constraints ( spelled out in detail below ) that must be obeyed by any dimer configuration that is dual to a spin configuration with periodic boundary conditions . if the global constraints are satisfied , one maps the new dimer configuration back to spin variables , to obtain an updated spin configuration which can differ from the original spin configuration by large nonlocal changes .
since the procedure explicitly satisfies detailed balance , one obtains in this way a valid algorithm for the spin model being studied . for the triangular lattice ising antiferromagnet , we have developed two strategies for constructing rejection - free updates of the generalized dimer model on the dual honeycomb lattice . as mentioned in the introduction , one of these generalizes readily to the generalized dice lattice dimer model which is dual to the frustrated kagome lattice ising model , while the other is specific to the generalized honeycomb lattice dimer model .
the strategy that generalizes readily to the dice lattice case is one in which we deliberately _ do not _ keep track of the local dimer configuration of the dual lattice at alternate steps of the worm construction in order to ensure that detailed balance can be satisfied without any final rejection step .
this is similar to the myopic worm construction developed earlier@xcite for a general class of link - current models@xcite , and used successfully on the dice lattice in earlier work on an interacting dimer model for the low temperature properties of certain high - spin kagome lattice antiferromagnets with strong easy - axis anisotropy.@xcite since this strategy involves being deliberately short - sighted at alternate steps of the construction , we dub this the `` myopic '' worm algorithm . below
we begin with a detailed description of how this works for the generalized dimer models on the honeycomb and dice lattices which are dual to the physics of the frustrated ising models on the triangular and kagome lattices .
the first step of the myopic worm construction on the dual dice lattice : a start site @xmath30 is chosen randomly from one of the six - coordinated sites on the dice lattice . the tail of the worm remains static at this start site until the worm construction is complete . in this first step , the head of the worm moves to one of the six neighbours of the start site with probability @xmath51 , regardless of the local dimer configuration .
the neighbour thus reached is our first vertex site @xmath32 .
viewed from the point of view of this vertex , the start site @xmath30 is the first entry site @xmath33 . , width=317 ]
the _ myopic _ step of the myopic worm algorithm on the dual dice lattice : after arriving at an exit site @xmath40 ( which , by construction , is always a six - coordinated site ) from a vertex site @xmath38 , the next vertex site @xmath48 is chosen to be one of the five _ other _ neighbours of @xmath40 with probability @xmath52 . viewed from this new vertex site @xmath48 , @xmath40 becomes the entry site @xmath50 .
, width=317 ] if the randomly chosen start site @xmath53 is touched by only one dimer , we move along that dimer to reach a new pivot site @xmath54 in the first step of the dep worm construction .
the start site @xmath53 becomes our first entry site @xmath55 .
, width=317 ] the pivot step of the dep worm construction when one arrives at a central pivot site @xmath56 from an entry site @xmath57 and there is only one dimer touching @xmath56 .
we pivot the dimer from the link @xmath58 to the link @xmath59 with probabilities determined by the corresponding elements of the probability table @xmath18 .
@xmath60 , which is the new overlap site , can either one of the two neighbours @xmath61 and @xmath62 of the central pivot site @xmath56 or the entry site @xmath57 from which we came to @xmath56 .
the central pivot site @xmath56 now becomes the new entry site @xmath63 from which this new overlap site @xmath60has been reached , and the next step is an overlap step .
on the dual honeycomb lattice , knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath47 and @xmath44 , @xmath45 and @xmath46 suffices to calculate @xmath18 when the interactions extend up to next - next - nearest neighbours on the triangular lattice . , width=317 ] on the honeycomb lattice this myopic worm algorithm consists of the following steps : we begin by choosing a random `` start site '' @xmath30 on the honeycomb lattice .
regardless of the local dimer configuration in the vicinity of this site , we move from the start site to one of the three neighbouring sites , with probability @xmath31 each ( fig .
[ start_myopic_triangular ] ) .
the neighbouring site reached in this way is our first `` vertex site '' @xmath32 . in our terminology , we have `` entered '' this vertex site from the start site @xmath30 .
therefore , the start site is the `` entry site '' @xmath33 for this vertex .
next , we choose one of the neighbours of @xmath32 as the `` exit site '' @xmath64 , via which we can exit this vertex .
when we arrive at vertex site @xmath32 from entry site @xmath33 , and leave this vertex site via exit site @xmath64 , we flip the dimer state of the dual links @xmath65 and @xmath66 . the choice of exit site @xmath64 via which we exit from a vertex site @xmath32 , given that we arrived at vertex site @xmath32 from a particular entry site @xmath33 , is probabilistic ( fig .
[ myopic_3by3 ] ) , with probabilities specified in a probability table @xmath41 whose structure we now discuss .
the pivot step of the dep worm construction when one arrives at a central pivot site @xmath56 from an entry site @xmath57 and there are three dimers touching @xmath56 : at this point , one has three options , with probabilities determined by corresponding entries of the probability table @xmath67 : we can choose to exit to one of the two neighbours @xmath61 or @xmath62 , or bounce back to the entry site @xmath57 .
if we choose to exit through either @xmath61 or @xmath62 , we move along the dimer connecting the central pivot site @xmath56 to this chosen exit which becomes our new pivot site @xmath68 , and delete the dimer on the link @xmath69 .
the third option is to move along the dimer connecting the central pivot site @xmath56 back to the entry site @xmath57 , and @xmath57 then becomes our new overlap site @xmath60 . on the dual honeycomb lattice , knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath47 and @xmath44 , @xmath45 and @xmath46 suffices to determine @xmath67 when the interactions extend upto next - next - nearest neighbours on the triangular lattice .
, width=317 ] the overlap step of the dep worm construction : one arrives at a central overlap site @xmath70 from an entry site @xmath57 . at this stage ,
one uses the probability table @xmath67 to choose one out of two options : if we choose to exit along the empty link ( in this case @xmath71 ) , we _ deposit _ a dimer on the empty link , and move along it making @xmath61 our new overlap site @xmath60 . the central overlap site
@xmath70 now becomes the new entry site @xmath63 from which we enter the new overlap site @xmath60 . on the other hand , we may choose to exit along the link @xmath72 to reach our new pivot site @xmath73 .
the central overlap site @xmath70 now becomes the new entry site @xmath63 , from which we enter the new pivot site @xmath68 . on the dual honeycomb lattice ,
knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath47 and @xmath44 , @xmath45 and @xmath46 suffices to determine the probability table @xmath67 , when the interactions extend upto next - next - nearest neighbours on the triangular lattice .
, width=317 ] for any vertex site @xmath74 encountered in our process , these probabilities are given by a probability table @xmath75 , where @xmath76 is the entry site from which we have entered the vertex and @xmath77 is the exit site we wish to leave from .
entries in this probability table are constrained by the requirement of local detailed balance . in order to state these constraints on @xmath78 in a way that makes subsequent analysis easy ,
we rewrite this table as a three - by - three matrix @xmath79 ( @xmath80 ) by choosing a standard convention to label the three neighbours of @xmath74 by integers running from one to three .
thus , if @xmath76 is the @xmath81 neighbour of @xmath74 and @xmath77 is the @xmath82 neighbour of @xmath74 according to this convention , we write @xmath83 .
we denote by @xmath84 the boltzmann weight of the dual dimer configuration before we flip the dimer states of dual links @xmath85 and @xmath86 . in the same way , @xmath87 , for each choice of @xmath77 ,
denotes the corresponding boltzmann weight after these flips are implemented .
as is usual for all worm algorithms , these weights for the intermediate configurations encountered during this myopic construction are obtained from the boltzmann weight of the generalized dimer model with the proviso that the `` infinite energy cost '' of violating the generalized dimer constraints at the start site and current site ( `` head '' and `` tail '' of the worm in worm algorithm parlance ) are ignored when keeping track of the weights of these intermediate configurations .
we choose the @xmath41 matrices to satisfy a local detailed balance condition that depends on these weights @xmath88 rewriting @xmath89 if @xmath76 is the @xmath81 neighbour of @xmath74 , and @xmath90 if @xmath77 is the @xmath82 neighbour of @xmath74 , we can write these detailed balance conditions in terms of the matrix @xmath79 and the weights @xmath91 ( with @xmath92 ) as @xmath93 as is usual in the analysis of such detailed balance constraints@xcite , we define the three - by - three matrix @xmath94 and note that the detailed balance condition is now simply the statement that @xmath95 is a symmetric matrix which satisfies the three constraints @xmath96 the first step of the dep worm construction when the randomly chosen start site @xmath53 is touched by three dimers : we move along any one of the three dimers with probability @xmath31 , to reach a new pivot site @xmath54 .
the start site @xmath53 becomes our first entry site @xmath55 .
, width=317 ] the pivot step of the dep worm construction when one arrives at a central pivot site @xmath56 from an entry site @xmath57 , and there are two dimers touching @xmath56 : we can exit to the neighbour @xmath76 , which is the neighbour not connected to the central pivot site @xmath56 by a dimer , or bounce back by exiting through the entry site @xmath57 .
if we choose to exit through the neighbour @xmath76 , we flip the dimer on the link @xmath58 to the link @xmath97 .
we then move along this dimer to reach our new overlap site @xmath98 .
if we choose to bounce back , the entry site @xmath57 becomes our new overlap site @xmath60 .
the central pivot site @xmath56 now becomes our new entry site @xmath63 in either case .
the probability table @xmath41 is used to determine which option is chosen . on the dual honeycomb lattice , knowledge of the local dimer configuration consisting of dimer states from @xmath42 to @xmath47 and @xmath44 , @xmath45 and @xmath46 suffices to calculate @xmath41 when the interactions extend upto next - next - nearest neighbours on the triangular lattice . , width=317 ] for interactions that extend up to next - next - nearest neighbours on the triangular lattice , the three weights @xmath99 that enter these equations differ from each other only due to factors that depend on the dimer state , @xmath100 of the the three links emanating from @xmath74 and the twelve dual links surrounding @xmath74 , whose dimer state has been denoted @xmath42 , @xmath101 @xmath47 in fig . [ myopic_3by3 ] .
this feature allows us to tabulate all possible local environments of @xmath74 , and analyze these constraint equations in advance to determine and tabulate the @xmath102 ( and thence determine @xmath103 ) in advance . in practice ,
if the weights permit it , we use the `` zero - bounce '' solution given in ref . and ref .
, else the `` one - bounce '' solution given there .
having reached the exit @xmath64 of the vertex @xmath32 in this manner , we now need to choose the next vertex @xmath104 which we will enter next from this site @xmath64 .
this is the myopic part of our procedure : _ this next vertex @xmath104 is randomly chosen to be one of the two other neighbours of @xmath64 ( other than the previous vertex @xmath32 ) with probability @xmath49 each_(fig . [ myopic_triangular ] ) . after making this choice
, @xmath64 becomes the entry site @xmath105 for this next vertex @xmath104 , and the previous probabilistic procedure is repeated at this next vertex @xmath104 in order to choose the next exit site @xmath106 from which we will exit @xmath104 . in this manner
, we go through a sequence of vertices until the exit site @xmath107 of the @xmath108 vertex equals the start site @xmath30 .
when this happens , one obtains a new dimer configuration which again has either one dimer touching each honeycomb site , or three dimers touching a honeycomb site .
this new dimer configuration can be accepted with probability one since our procedure builds in detailed balance with respect to the boltzmann weight of the generalized dimer model .
it is straightforward to prove this explicitly using the notation we have developed above . to this end
, we first note that the forward probability for constructing a particular worm to go from an initial configuration @xmath109 to a final configuration @xmath110 takes on the product form @xmath111 while the reverse probability takes the form @xmath112 as noted earlier , while the weights @xmath113 that appear in the intermediate steps of the construction are computed ignoring the violation of the generalized dimer constraints at two sites , the initial and final weights @xmath114 and @xmath115 have no such caveats associated with them .
indeed , we have @xmath116 the physical boltzmann weight of the initial configuration , while @xmath117 the physical boltzmann weight of the final configuration . now , since our choice of transition probabilities obeys @xmath118 for all @xmath119 , and since @xmath120 for all @xmath121 , we may write the following chain of equalities @xmath122 thus , our procedure explicitly obeys detailed balance , and this myopic worm construction provides a rejection - free update scheme that can effect large changes in the configuration of a generalized honeycomb lattice dimer model with one or three dimers touching each honeycomb site . the frequency of configurations accessed by a monte carlo simulation consisting of @xmath123 monte carlo steps ( as defined in sec . [ performance ] ) for a @xmath124 triangular lattice ising model with interactions extending to next - next - nearest neighbours
there are 512 possible unique configurations .
the frequencies measured for the dep worm algorithm and the myopic worm algorithm are seen to agree well with the predictions of equilibrium gibbs - boltzmann statistics , thus establishing the validity of both algorithms.,width=317 ] to translate back into spin language , we need to take care of one additional subtlety : although the procedure outlined above gives us a rejection - free nonlocal update for the generalized dimer model with boltzmann weight inherited from the original spin system , we can not translate this directly into a rejection - free nonlocal update for the original spin system since we are working on a torus with periodic boundary conditions for the spin system .
the reason has to do with the fact that the periodic boundary conditions of the spin system translate to a pair of global constraints : in every valid dimer configuration obtained from a spin configuration with periodic boundary conditions , the number of empty links crossed by a path looping around the torus along @xmath36 or @xmath37 must be even , since the absence of a dimer on a dual link perpendicular to a given bond of the spin system implies that the spins connected by that bond are antiparallel .
this corresponds to constraints on the global winding numbers of the corresponding dimer model ( see ref . for a definition specific to the honeycomb lattice dimer model ) , which must be enforced by any monte carlo procedure . note that these constraints are on the parity of these winding numbers , which are in any case only defined modulo @xmath125 unless one is at @xmath3 . therefore , to convert this rejection - free myopic worm update procedure for dimers into a valid update scheme for the original spin system , we test the winding numbers ( modulo @xmath125 ) of the new dimer configuration to see if it satisfies these two global constraints .
if the answer is yes , we translate the new dimer configuration back into spin language by choosing the spin at the origin to be up or down with probability @xmath49 and reconstructing the remainder of the spin configuration from the positions of the dimers . if , however , the new dimer configuration is in an illegal winding sector , we repeat the previous spin configuration in our monte carlo chain .
this procedure generalizes readily to the kagome lattice ising antiferromagnet with interactions extending up to next - next - nearest neighbour spins . since most of the required generalizations are self - evident , we merely point out some of the key differences here .
our myopic worm update procedure now begins with a randomly chosen six - coordinated site as the start site @xmath30 . with probabilities
@xmath51 each , we choose one of its neighbours as the first vertex site @xmath32 ( fig . [ start_myopic_kagome ] ) .
the start site @xmath30 thus becomes the entry site @xmath33 from which we enter the first vertex @xmath32 .
the choice of the first exit site @xmath64 via which we exit the first vertex is again dictated by a three - by - three probability table(fig .
[ myopic_3by3 ] ) . for any vertex @xmath74
, this probability table is determined by solving detailed balance equations completely analogous to the ones displayed earlier for the honeycomb lattice case . in the dice lattice case ,
the three weights @xmath91 depend on the dimer states @xmath42 , @xmath101 @xmath43 of the six dual links shown in fig
. [ myopic_3by3 ] , and on the dimer states @xmath100 of the three links emanating from @xmath74 .
therefore , we are again in a position to solve these equations for all possible local environments of @xmath74 and tabulate these solutions for repeated use during the worm construction . as in the honeycomb case , having arrived at @xmath64 , we choose the next vertex @xmath104 in a myopic manner : without regard to the local dimer configuration , we randomly pick , with probability @xmath52 each , one of the other neighbours ( other than @xmath32 ) of @xmath64 as the next vertex @xmath104 ( fig .
[ myopic_kagome ] ) .
@xmath64 now becomes the entry site @xmath105 from which we enter this second vertex @xmath104 .
the exit @xmath106 is again chosen from the pre - tabulated probability table , and the process continues until the @xmath108 exit @xmath107 equals then start site @xmath30 . clearly , our earlier proof of detailed balance goes through unchanged , and this myopic worm construction again gives a rejection - free way of updating the dual dimer model in accordance with detailed balance . to translate this into an update scheme for the original spin model
, we must again check that the new dimer configuration is in a legal winding sector , and if the new configuration is in an illegal winding sector , we must repeat the original spin configuration in our monte carlo chain . the frequency of configurations accessed by a monte carlo simulation consisting of @xmath126 monte carlo steps ( as defined in sec .
[ performance ] ) for a @xmath127 kagome lattice ising model with interactions extending to next - next - nearest neighbours .
there are 4096 possible unique configurations .
the frequencies measured for the myopic worm algorithm are seen to agree well with the predictions of equilibrium gibbs - boltzmann statistics , thus establishing the validity of the algorithm .
, width=317 ] the other strategy we have developed is specific to the honeycomb lattice dimer model that is dual to the triangular lattice ising antiferromagnet .
since it involves deposition , evaporation , and pivoting of dimers , we dub this the dep worm algorithm .
the dep worm construction begins by choosing a random start site @xmath53 .
the subsequent worm construction consists of so - called `` overlap steps '' and `` pivot steps '' .
pivot steps are carried out when one reaches a `` central '' `` pivot site '' from a neighbouring `` entry site '' , while overlap steps are carried out when one reaches a central `` overlap site '' from a neighbouring entry site .
details of some of the subsequent pivot steps in this construction depend on whether the randomly chosen start site @xmath53 is touched by three dimers or by one dimer , _
i.e. _ if the corresponding triangle is defective or minimally frustrated
. therefore we describe these two branches of the procedure separately , but use a unified notation so as to avoid repetition of the aspects that do not depend on the branch chosen .
let us first consider the case when the randomly chosen start site @xmath53 is touched by exactly one dimer(fig .
[ start_branch1 ] ) . in this case
, we move along the dimer touching @xmath53 to its other end . the site at the other end of this dimer becomes our first central site @xmath54 , at which we must now employ a pivot step with @xmath54 as the first pivot site .
for the purposes of this first pivot step , the start site @xmath53 becomes the entry site @xmath55 from which we have arrived at this pivot site @xmath54 by walking along this dimer . before proceeding further ,
it is useful to elucidate the nature of a general pivot move encountered in our algorithm : in a pivot step , after one arrives at the central pivot site @xmath128 from an entry site @xmath57 ( as we will see below , @xmath57 could be the previous overlap site @xmath129 , or a previous pivot site @xmath130 ) by moving along a dimer connecting @xmath57 to @xmath128 , the subsequent protocol depends on whether there is exactly one dimer ( fig . [ loop ] ) touching @xmath128 or three ( fig .
[ kill ] ) . in the first case ,
one pivots the dimer touching @xmath128 , so that it now lies on link @xmath131 instead of link @xmath132(fig . [ loop ] ) . here
, @xmath133 is one of the neighbours of @xmath128 , chosen using the element @xmath134 of a three - by - three probability table @xmath135 ( where @xmath136 and @xmath137 range over the three neighbours of the central site @xmath128 , and the full structure of this table is specified at the end of this discussion ) .
note that in some cases , it is possible for @xmath138 with nonzero probability , if the corresponding diagonal entry of the table is nonzero .
after this is done , the next step in the construction will be an overlap step , with @xmath133 being the central overlap site and @xmath139 playing the role of the new entry site @xmath140 from which we have arrived at this central overlap site .
the structure of a general overlap step is specified below , after describing the pivot move in the second case , _
i.e. _ with three dimers touching the central pivot site . if the central pivot site @xmath128 in a pivot step has three dimers connecting it ( fig .
[ kill ] ) to its three neighbours @xmath61 , @xmath62 , and @xmath57 ( where @xmath57 is the entry site from which we arrived at the central pivot site @xmath128 ) , we choose one out of three alternatives using a different probability table @xmath141 , where @xmath136 and @xmath142 range over all neighbours of a central site @xmath143 and @xmath144 is a particular privileged neighbour of @xmath143 ( in the case being described here , @xmath145 and @xmath146 ) : with probabilities @xmath147 and @xmath148 drawn respectively from this table , we may delete the dimer on link @xmath132 and reach either @xmath61 or @xmath62 , and the next step would then be a pivot step , with the neighbour thus reached now playing the role of the new central pivot site @xmath149 and @xmath128 playing the role of the new entry site @xmath140 from which we have reached this new central pivot site . on the other hand , we may `` bounce '' with probability @xmath150 , _ i.e. _ we simply return from @xmath128 to @xmath57 without deleting any of the three dimers touching @xmath128 ; in this case , the next step will be an overlap step , with @xmath57 as the new central overlap site @xmath133 , and @xmath128 will play the role of the new entry site @xmath140 from which we have reached this new overlap site ( fig . [ kill ] ) .
note that elements of this table @xmath141 with @xmath151 never play any role in the choices made at this kind of pivot step .
as we will see below , these elements of the table in fact determine the choices made at a general overlap step in a way that preserves local detailed balance .
returning to our construction , if the central pivot site @xmath54 was of the second type and we did not bounce , we would reach a new central pivot site @xmath152 ( with @xmath54 now becoming the entry site @xmath153 from which we reach this new pivot site ) , and we would perform another pivot step as described above . if on the other hand , the central pivot site @xmath54 was of this first type or if it was of the second type and we bounced , the next step will be an overlap step with a new central overlap site @xmath152 .
since we would have reached @xmath152 by moving along a dimer connecting it to @xmath54 , @xmath54 will play the role of the new entry site @xmath153 for this overlap step ( in the bounce case , @xmath154 ) .
having reached the central overlap site @xmath152 from entry site @xmath153 in this way , we must employ an overlap step . before proceeding with our construction ,
let us first elucidate the structure of choices at an overlap step after we have reached a central overlap site @xmath70 from an entry site @xmath57 .
@xmath57 could be the previous pivot site @xmath155 if the previous step had been a pivot step ( as in the example above ) or it could be a previous overlap site @xmath156 if the previous step had also been an overlap step ( we will see below that this is also possible ) . in either case , at a general overlap step , one arrives at the overlap site @xmath70 from entry site @xmath57 along one of the two dimers touching @xmath70 .
thus , one neighbour of @xmath70 , suggestively labeled @xmath133 , is _ not _ connected to the central overlap site @xmath70 by a dimer , while the other two neighbours are connected to @xmath70 by dimers .
one of the latter pair of neighbours is of course the entry site @xmath57 from which we arrived at @xmath70 , while we suggestively label the other as @xmath149 .
at such an overlap step , one always has two options to choose from , whose probabilities are given as follows by entries of the probability table @xmath67 introduced earlier(fig [ make ] ) : one option is to deposit , with probability @xmath157 , an additional dimer on the originally empty link @xmath158 emanating from @xmath159 .
if we do this , @xmath133 becomes the new overlap site , which we have entered from @xmath70 , which becomes the new entry site @xmath140 , and the next step will again be an overlap step .
the second option , chosen with probability @xmath160 , is that we move along the second dimer touching @xmath159 to the other neighbour @xmath149 , which is connected to @xmath159 by this second dimer .
if we do this , @xmath149 becomes the new pivot site , which we enter from site @xmath70 , which becomes the new entry site @xmath140 , and the next step will be a pivot step . as we will see below ,
the fact that the table @xmath67 that fixes the probabilities for choosing between these two options is the same as the one used in a pivot step ( when the pivot site has three dimers touching it ) is crucial in formulating and satisfying local detailed balance conditions that guarantees a rejection - free worm update . returning again to our construction , we employ this procedure to carry out an overlap step when we reach the overlap site @xmath161 from entry site @xmath153 .
clearly this process continues until we encounter the start site @xmath53 as the new overlap site in the course of our worm construction .
when this happens , we obtain a new dimer configuration that satisfies the generalized dimer constraint that each site be touched by one or three dimers . the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the dep worm algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at three temperatures at which the system is in the power - law ordered critical phase .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding temperatures .
, width=317 ] the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the myopic worm algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at three temperatures at which the system is in the power - law ordered critical phase .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding temperatures .
, width=317 ] the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the dep worm algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at six values of @xmath165 at which the system is in the power - law ordered critical phase in the zero temperature limit @xmath166 .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding points in the zero temperature phase diagram .
, width=317 ] let us now consider the case when the randomly chosen start site @xmath53 is touched by three dimers . in this case ,
we move along one of the three dimers touching @xmath53 to its other end ( with probability @xmath31 each ) , so that the site at the other end becomes the central pivot site @xmath54 for a pivot step , and the start site @xmath53 becomes the entry site @xmath55 from which we have entered this central pivot site ( fig [ start_branch2 ] ) .
we now implement the protocol for a pivot step ( as described in _
branch i _ ) to reach a new central site @xmath152 .
if the next step turns out to be a pivot step , @xmath152 plays the role of a central pivot site , whereas it becomes the central overlap site if the next step is an overlap step . in either case
, @xmath54 becomes the entry site @xmath153 for this next step . in this manner
, we continue until we reach the start site @xmath53 as the new central overlap site @xmath167 for an overlap step .
when this happens , the worm construction ends after these @xmath168 steps , since the start site again has three dimers touching it , and we thus obtain a new dimer configuration that satisfies the generalized dimer constraint that each site be touched by one or three dimers . the only additional feature introduced in _
branch ii _ is that one could in principle reach the start site @xmath53 as the central pivot site of some intermediate pivot step @xmath169(fig .
[ 2by2 ] ) . in this case , the intermediate configuration reached at this @xmath170 step is not a legal one ( since it still has two dimers touching @xmath53 ) , and we need to continue with the worm construction .
this is done using a special `` two - by - two '' pivot step(fig . [ 2by2 ] ) . in this two - by - two pivot step ,
one arrives at the two - by - two pivot site ( which will always be the start site in our construction ) @xmath56 from an entry site @xmath57 ( as in all other steps , @xmath57 could be a previous central overlap site @xmath129 or the previous central pivot site @xmath130 ) by moving along a dimer connecting @xmath57 to @xmath56 .
unlike the usual pivot step , at which there is only one dimer touching the central pivot site , @xmath56 has a second dimer touching it , which connects @xmath56 to another neighbour @xmath171 .
thus , unlike the usual pivot step , there is just one neighbour of @xmath56 , suggestively labeled @xmath133 , which is not connected to @xmath56 by a dimer when one arrives at @xmath56 to implement this step .
therefore , our only options are to rotate the dimer which was on link @xmath172 , to now lie on link @xmath173 , or to bounce . the probabilities for these two choices
are determined by a probability table @xmath174 . here , @xmath136 and @xmath142
are both constrained to not equal @xmath171 , making @xmath174 a two - by - two matrix . in either case ,
@xmath133 chosen in one of these two ways becomes the new central overlap site of the next step , which must be an overlap step , and the process continues .
this new configuration thus obtained upon completing the worm construction initiated either using _ branch i _ or _ branch ii _ can now be accepted with probability one if the probabilities with which we carried out each of the intermediate pivot steps and overlap steps obeyed local detailed balance .
local detailed balance at a pivot step in which the pivot site is touched by one dimer requires that the probability table @xmath175 obeys the conditions @xmath176 where the @xmath177 is the boltzmann weight of the dimer configuration in which the link @xmath178 connecting @xmath56 to one of its neighbours @xmath76 is occupied by a dimer and the other two links emanating from @xmath56 are empty .
as in all worm constructions , these weights are computed ignoring the fact that the generalized dimer constraint ( that each site be touched by exactly one or three dimers ) is violated at two sites on the lattice .
these conditions again form a three - by - three set of constraint equations of the type discussed in ref . and ref .
, allowing us to analyze these constraints and tabulate solutions in advance for all cases that can be encountered .
if the weights permit it , we use the `` zero - bounce '' solution given in ref . and ref . , else the `` one - bounce '' solution given there
local detailed balance at a two - by - two pivot step in which the pivot site is touched by two dimers requires that the probability table @xmath174 obeys the conditions @xmath179 where the @xmath180 is the boltzmann weights of the dimer configurations in which the links @xmath181 and @xmath182 are covered by dimers and the third link is empty . as always , these weights are computed ignoring the fact that the dimer constraints are violated at two sites on the dual lattice . in practice
, we tabulate all possible local environments that can arise in such an update step , and use metropolis probabilities to tabulate in advance the corresponding entries of @xmath183 .
the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the myopic worm algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at six values of @xmath165 at which the system is in the power - law ordered critical phase in the zero temperature limit @xmath166 .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding points in the zero temperature phase diagram .
, width=317 ] finally , the constraints imposed by local detailed balance at an overlap step are essentially intertwined with the local detailed balance constraints that must be enforced at a pivot step when the pivot site has three dimers touching it .
this is because the deletion of a dimer at such a pivot step is the `` time - reversed '' counterpart of the process by which an additional dimer is deposited at an overlap step .
indeed , this is why we have been careful in our discussion above to draw the probabilities at the pivot step from the same table @xmath67 as the probabilities that govern the choices to be made at an overlap step . the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the myopic worm algorithm on the kagome lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at three temperatures at which the system is in the power - law ordered critical phase .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding temperatures .
, width=317 ] we use a different probability table @xmath141 ( where @xmath136 and @xmath142 range over all neighbours of a central site @xmath143 and @xmath144 is a particular privileged neighbour of @xmath143 ) to decide on the next course of action : we now describe the structure of the table @xmath141 . here
, @xmath143 is the central site , which would be the current pivot site in a pivot step with three dimers touching the pivot site , or the current overlap site in an overlap step .
@xmath144 is a `` privileged neighbour '' of @xmath143 ; in a pivot step , @xmath144 is the entry site from which we enter the pivot site @xmath143 , while in an overlap step , it is the unique neighbour of @xmath143 that is not connected to @xmath143 by a dimer .
clearly , local detailed balance imposes the following constraints on this probability table @xmath67 : @xmath184 here both @xmath136 and @xmath142 can be either the site @xmath144 or the two other neighbours @xmath61 and @xmath62 of the central site @xmath143 .
@xmath185 denotes the weight of the configuration with both links @xmath186 and @xmath187 covered by a dimer and the link @xmath188 unoccupied by a dimer . on the other hand
, @xmath189 denotes the weight of the configuration in which all three links @xmath186 , @xmath187 and @xmath190 are covered by dimers . as before ,
these weights are computed ignoring the fact that the generalized dimer constraint ( that each site be touched by exactly one or three dimers ) is violated at two sites on the lattice .
choices for the tables @xmath18 and @xmath67 consistent with these local detailed balance constraints , can be computed using the same strategy described in our construction of the myopic worm update .
again , the weights that enter these constraints on @xmath67 ( @xmath18 ) depend only on the dimer states @xmath44 , @xmath45 , and @xmath46 of the three links emanating from the central site @xmath143 ( pivot site @xmath56 ) , and the dimer states @xmath42 , @xmath101 @xmath47 of the twelve links surrounding this site , allowing us to tabulate in advance all possible local environments and the corresponding solutions for @xmath67 and @xmath18 . the formal proof of detailed balance uses these local detailed balance constraints to construct a chain of equalities relating the probabilities for an update step and its time - reversed counterpart in exactly the same way as the proof given in the previous discussion of the myopic worm update . therefore , we do not repeat it here for the present case . the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the metropolis algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at three temperatures at which the system is in the power - law ordered critical phase .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding temperatures .
, width=317 ] the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the metropolis algorithm on the triangular lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at six values of @xmath165 at which the system is in the power - law ordered critical phase in the zero temperature limit @xmath166 .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding points in the zero temperature phase diagram .
, width=317 ] the lattice size @xmath162 dependence of autocorrelation time of the three sublattice order parameter @xmath163 in monte carlo simulations using the metropolis algorithm on the kagome lattice with nearest neighbour antiferromagnetic and next - nearest neighbour ferromagnetic interactions .
the dynamical exponent @xmath11 is extracted by fitting to the functional form @xmath164 at three temperatures at which the system is in the power - law ordered critical phase .
also shown is the anomalous exponent @xmath12 of the power law three - sublattice order at the corresponding temperatures .
, width=317 ] the dynamical exponent @xmath191 for the three - sublattice order parameter in simulations employing the dep and myopic worm algorithms depends in a universal way , independent of the lattice as well as the details of the worm construction procedure , on the corresponding equilibrium anomalous exponent @xmath12 .
however , @xmath192 appears to be nearly constant in the nonzero temperature power - law ordered phase , while the corresponding function in the zero temperature limit is clearly different.,width=317 ] dynamical exponent @xmath191 for the three - sublattice order parameter in simulations employing the metropolis algorithm .
, width=317 ]
as mentioned earlier , the next - nearest and next - next - nearest neighbour interactions induce three - sublattice long - range order of the ising spins on both triangular @xcite and kagome @xcite lattices .
this order melts via a two - step transition with an intermediate power - law ordered critical phase .
the correlation function of the three - sublattice order in this phase decays as a power law with a temperature dependent exponent @xmath193 . here
, we use this power - law ordered intermediate phase associated with two - step melting of three - sublattice order as a challenging test bed for our algorithms . in this regime on the triangular lattice , we compare the performance of the dep worm algorithm and the myopic worm algorithm with that of the metropolis algorithm . similarly ,
we compare the performance of the myopic worm algorithm on the kagome lattice with that of the single spin flip metropolis algorithm in the same critical phase . on the triangular lattice we use @xmath7,@xmath194 and @xmath195 , while for the kagome lattice system
, we use @xmath7,@xmath196 and @xmath195 .
monte carlo simulations are performed on both lattices for three values of @xmath41 corresponding to three values of the anomalous exponent @xmath12 . on the triangular lattice in the limiting case of @xmath166 with @xmath197 , a power - law phase@xcite
is also established for a range of @xmath9 .
this @xmath3 phase is characterized by a @xmath9-dependent @xmath198 . with this in mind , we also study this @xmath166 limit on the triangular lattice , with @xmath7 and @xmath195 , for six values of @xmath199 corresponding to six values of @xmath12 in this @xmath3 power - law ordered phase . as is conventional ,
we characterize the three - sublattice order in terms of the complex three - sublattice order parameter @xmath200 .
additionally , we also monitor the ferromagnetic order parameter @xmath0 .
these are defined as : @xmath201 and @xmath202 on the triangular lattice and @xmath203 and @xmath204 on the kagome lattice , where @xmath205 is used to label the sites as shown in fig . [ lattice ] and @xmath206 $ ] , labels the three basis sites in each unit cell of the kagome lattice ( fig . [ lattice ] ) . in order to meaningfully compare the algorithms ,
we need a consistent definition of one monte carlo step .
this is achieved as follows : for the dep and myopic worm algorithms , we first compute the average number of sites visited by a complete worm of the algorithm at a given point in parameter space .
we then adjust the number of worms in one monte carlo step ( mcs ) so that the average number of sites visited in one mcs equals the number of sites in the lattice . for the metropolis algorithm
, we simply fix the number of attempted spin flips in a step to be equal to the number of sites on the lattice , and this defines one mcs for the metropolis algorithm . to validate the worm algorithms , we first study the frequency table of the accessed configurations in long runs on a @xmath124 triangular lattice with the dep and myopic worm algorithms and a @xmath127 kagome lattice with the myopic worm algorithm . as is clear from our results , the measured frequencies are perfectly predicted by theoretical expectations based on the boltzmann - gibbs probabilities of different configurations ( see fig . [ t_freq_allj ] for the triangular lattice and fig .
[ k_freq_allj ] for the kagome lattice ) . with all algorithms thus performing an equivalent amount of work in one mcs , we can compare the performance of these algorithms by measuring the autocorrelation function ( defined below ) of the order parameter . given the presence of a critical phase with power - law three - sublattice order , it is natural to test the performance of our algorithms by analyzing the @xmath162 dependence of the autocorrelation time .
more precisely , for a given observable @xmath207 , whose value at the @xmath208 mcs is depicted as @xmath209 , we define the normalized autocorrelation function @xmath210 in the standard way : @xmath211 where @xmath212 implies averaging over the monte carlo run .
we extract autocorrelation times @xmath213 and @xmath214 by fitting @xmath215 for the dep and myopic worm algorithms as well as the metropolis algorithm at various system sizes @xmath162 to exponential relaxation functions .
we plot these autocorrelation times as a function of @xmath162 and fit these curves to the power law functional form @xmath164 in order to extract the corresponding monte carlo dynamical exponent @xmath11 . figs .
[ t_gen_ac_psi_j1_j2 ] , [ t_myo_ac_psi_j1_j2 ] , [ t_gen_ac_psi_onlyj2 ] , [ t_myo_ac_psi_onlyj2 ] , [ k_myo_ac_psi_j1_j2 ] show such power law fits to the @xmath162 dependence of the autocorrelation time @xmath213 .
these fits provide us our estimates for the dynamical exponent @xmath191 .
[ t_met_ac_psi_j1_j2 ] , [ t_met_ac_psi_onlyj2 ] and [ k_met_ac_psi_j1_j2 ] show the corresponding plots for the metropolis algorithm . the @xmath12 dependence of the dynamical exponent @xmath191 for the dep and myopic worm algorithms is shown in a comparative plot in fig .
[ etavsz_psi ] . from these results , it is clear that @xmath191 is quasi - universal in the following sense : in the power - law ordered phase at @xmath3 on the triangular lattice , @xmath191 is independent of the actual details of the worm construction protocol , and data from the dep algorithm and the myopic algorithm together define a @xmath3 functional form @xmath216 .
the @xmath4 results are also quasi - universal in a similar sense : data from the myopic algorithm on the kagome and triangular lattices together define a nearly - constant function @xmath216 .
the corresponding @xmath12 dependence of @xmath191 for the metropolis algorithm is shown in fig .
[ etavsz_psi_metro ] in both ( @xmath4 and @xmath3 ) power - law ordered phases . comparing this to our results for @xmath216
, we see that both the dep and myopic worm algorithms outperform the metropolis update scheme by a very wide margin .
corresponding results for the autocorrelation times of @xmath217 are detailed in the supplemental material .
our @xmath166 results for @xmath191 can also be compared to the findings of ref . ,
in which the efficiency of a plaquette - based generalization of the wolff cluster algorithm was studied in the context of the low temperature limit of the nearest neighbour triangular lattice ising antiferromagnet .
comparing to these results , we find that our approach outperforms this earlier algorithm even in this simple case .
additionally , this earlier approach does not generalize in an obvious way to systems with further neighbour couplings , while the approach described here is designed to incorporate such further - neighbour couplings in a straightforward way .
it is natural to wonder if the dual worm strategies introduced here could be ported to quantum monte carlo simulations of frustrated models .
unfortunately , no straightforward generalization of this type appears possible .
the difficulty is that a worm which alters the state on a given `` time - slice '' of the quantum monte carlo configuration also introduces unphysical off - diagonal terms in the hamiltonian , which correspond to ring - exchanges over large loops ( in the language of the dual quantum dimer model ) . in this context , it is worth nothing that the cluster construction strategy developed recently for frustrated transverse field ising models in ref .
reduces , in the limit of vanishingly small transverse field , to a version of the kandel - ben av - domany plaquette - percolation approach .
thus , generalizations to quantum systems appear to be more natural within that framework , rather than in the dual worm approach used here .
finally , we note that the quasi - universal behaviour of the dynamical exponents @xmath191 and @xmath218 is very suggestive , throwing up the possibility that the statistics of worms constructed by these algorithms is universally determined by the long - wavelength properties of the underlying equilibrium ensemble .
we hope to return to this in more detail in a future publication .
we thank d. dhar for useful discussions . our computational work relied exclusively on the computational resources of the department of theoretical physics at the tata institute of fundamental research ( tifr ) . | we report on the development of two dual worm constructions that lead to cluster algorithms for efficient and ergodic monte carlo simulations of frustrated ising models with arbitrary two - spin interactions that extend up to third - neighbours on the triangular lattice .
one of these algorithms generalizes readily to other frustrated systems , such as ising antiferromagnets on the kagome lattice with further neighbour couplings .
we characterize the performance of both these algorithms in a challenging regime with power - law correlations at finite wavevector . |
because stars are well - spaced , even the inner parts of galaxies can appear faint , comparable in surface brightness to the darkest of night skies . not surprisingly , then , when we start exploring the full six - dimensional phase space of velocities and positions that stars occupy in galaxies , the signal per phase space element becomes almost vanishingly small . in the case of the milky way ,
for example , the fascinating stellar streams in the halo @xcite have widths of only @xmath0 , and velocity dispersions of maybe @xmath1 .
if we were to divide the galaxy , with a linear dimension of @xmath2 and velocities spanning @xmath3 , into resolution elements this small , we would end up with @xmath4 of them , many more than there are stars in the milky way , so the average occupation number per element would be very small indeed .
thus , if we are to study the detailed kinematics of galaxies , we need to do two things .
first , we must be able to detect stellar populations at extremely low phase densities , which essentially means detecting individual stars .
second , we need to come up with ways to combine the information from these individual detections in order to bring together enough signal to say anything about the properties of the galaxy , and hence reconstruct its detailed dynamical structure in order to learn about its evolution .
planetary nebulae ( pne ) offer an ideal tracer of such faint stellar populations , as they are readily individually detectable from their emission lines , even in quite distant galaxies , and the doppler shifts in these lines can be used to measure their line - of - sight velocities .
to make such measurements efficiently in a single step , we have designed and built a customized instrument , the planetary nebula spectrograph or pn.s @xcite . in this paper , we present examples from a couple of the projects that we have been undertaking with this instrument , to illustrates the principles of investigating galaxies in the ultimate low - density regime of phase space .
when one traces edge - on stellar disks away from the plane out to very low surface brightnesses , an excess of light above the extrapolation of the normal disk population is frequently found @xcite .
the nature of such `` thick disks '' remains highly controversial : they could just be the extreme tail of the normal disk population , or they could represent some more dramatic event , such as the debris from a merger .
one exciting new clue to help distinguish between these possibilities was uncovered when it was found that the thick disk in the edge - on system fgc 227 appears to be rotating in the opposite direction to the normal thin - disk population , indicating that it would have to be a distinct component such as might be formed in a merger @xcite .
however , the surface brightness of this component is so faint that , even with ten hours of integration on an 8-metre telescope , conventional spectroscopy had such large error bars that the data were also consistent with co - rotating thin and thick disks .
[ fig : ngc891 ] as figure [ fig : ngc891 ] illustrates , we can do a great deal better with a rather shorter integration on a 4-metre telescope using pn.s .
the upper panel shows the number counts of pne detected in the edge - on spiral ngc 891 as a function of distance from its plane .
the lines show a possible decomposition of the disk into thin and thick exponential components @xcite .
the data drop below the model close to the plane , as pne are difficult to detect against the high surface brightness of the galaxy in this region , and some pne will also be lost in ngc 891 s strong dust lane . however , in the region we are interested in away from the plane , the pn number counts trace the light distribution very well , underlining their nature as generic tracers of the stellar population .
further , we are clearly detecting them in significant numbers out to beyond the transition where the thick disk becomes the dominant component .
the lower panel in this figure shows the observed mean rotation speed for these pne as a function of distance from the plane .
there is plenty of signal to distinguish between co- and counter - rotating thin and thick disk components , and in this system the two are clearly rotating in the same direction . indeed , we can go further and fit a simple single - component kinematic model to the data .
this heuristic model is of the form @xmath5 where @xmath6 provides the factor that allows for the line - of - sight projection effects that reduce the line - of - sight velocity below the mean rotational velocity @xcite , while @xmath7 parameterizes all the terms related to the shape of the velocity ellipsoid in the asymmetric drift equation @xcite ; here it is assumed to be a constant .
the circular speed of the galaxy , @xmath8 , can be inferred from gas kinematics , while the vertical velocity dispersion , @xmath9 , is estimated by solving the jeans equation for a self - gravitating sheet whose density distribution is given by the double exponential model in the upper panel of figure [ fig : ngc891 ] .
clearly , the data are essentially consistent with this simple single - component model : there is no kinematic evidence that the thick disk in ngc 891 is in any way a distinct entity .
as a further illustration of the way in which faint kinematic components can be detected , figure [ fig : m31 ] shows the results of an analysis of the 2615 pne that were found using pn.s in a survey of m31 @xcite . in an inclined disk system like this , the phase space is essentially reduced to four dimensions , two spatial and two velocity .
we can measure three of these components , two spatial and the line - of - sight velocity , so each pn has only one unknown coordinate .
thus , while one can not measure the energy and angular momentum of any single pn , there is only a one - dimensional family of possibilities , as illustrated by the line sections in the figure .
if group of pne share a common orbit ( and hence energy and angular momentum ) , then their lines will all intersect at a single point , creating a local excess in this plane .
such an excess of crossing lines is , indeed , found in this figure , at the point indicated by the box .
the location of the orbit to which it corresponds does not appear in any way special in an optical image of m31 , but , as the inset image in the figure shows , mid - infrared data reveal a bright ring of emission at this radius , indicating that it lies at one of the major orbital resonances .
thus , either we have detected the population of stars born in this ring of star formation , or , more likely since pne come from old stars , we are picking out the relatively small population of objects that have become dynamically trapped at this resonance .
whatever their origin , these data again illustrate the power of pn.s to identify kinematic components that are at far too low a density in phase space to be detectable by more conventional means .
k. gordon , et al . , `` the multiwavelength view of m31 including new spitzer / mips infrared images '' in _ bulletin of the american astronomical society _ , 2004 , vol .
36 of _ bulletin of the american astronomical society _ , 1579 . | galaxies are faint enough when one observes just their light distributions , but in studying their full dynamical structure the stars are spread over the six dimensions of phase space rather than just the three spatial dimensions , making their densities very low indeed .
this low signal is unfortunate , as stellar dynamics hold important clues to these systems life histories , and the issue is compounded by the fact that the most interesting information comes from the faintest outer parts of galaxies , where dynamical timescales ( and hence memories of past history ) are longest . to extract this information ,
we have constructed a special - purpose instrument , the planetary nebula spectrograph , which observes planetary nebulae as kinematic tracers of the stellar population , and allows one to study the stellar dynamics of galaxies down to extremely low surface brightnesses .
here , we present results from this instrument that illustrate how it can uncover the nature of low surface - brightness features such as thick disks by studying their kinematics , and trace faint kinematic populations that are photometrically undetectable .
address = school of physics & astronomy , university of nottingham , uk , [email protected] address = http://www.astro.rug.nl/~pns/ |
it has been recently pointed out @xcite that the radiative corrections to the masses of the higgs bosons in the framework of the mssm @xcite , @xcite , @xcite can be relatively large .
the leading correction to the effective potential comes from the top quark and squark loops , being proportional to the top yukawa coupling , which is considered to be big due to the heaviness of the top quark .
the other corrections happen to be smaller , though in some cases their effect is not negligible as well .
the net effect of the radiative corrections is to increase the masses of the higgs bosons .
this increase may be very significant for the future searches , since it can achieve several dozen gev , implying that the higgs mass could exceed the @xmath0 boson mass .
considering the tree - level higgs potential one finds out that the value of the lightest higgs boson mass is restricted by the inequality @xmath1 this strict limit is , however , violated by the radiative corrections .
the radiative corrections to the supersymmetric boson masses proceed > from the one - loop effective potential @xmath2 where @xmath3 denotes the conventional supertrace and @xmath4 is the scale at which all the couplings in the tree - level potential are renormalized .
@xmath5 are the field dependent masses of all the possible particles running through the loops . in what follows
we limit ourselves with the top and stop contributions as the main ones .
as far as the radiative corrections appear to be large achieving 30 % one can wonder about the values of the higher order contributions .
since according to eq.([3 ] ) they have the @xmath6 form one is expecting to have @xmath7 at the second loop , @xmath8 at the third loop , etc . being essential these logs have to be summed giving considerable change of the results .
indeed this happens in the simplest case of the @xmath9 model considered in the pioneering paper by coleman and weinberg @xcite .
the summation of the leading logs to the effective potential changed the situation qualitatively leading to the disappearance of a non - trivial minimum arising at the one - loop level .
the summation procedure can be naturally done with the help of the renormalization group technique , which we are going to apply to our particular case .
let us remind the expression for the one - loop effective higgs potential in the mssm which takes into account the radiative corrections due to the top quark and squark loops . for the neutral higgses
it has been calculated in ref.@xcite and has the form @xmath10 @xmath11-v_0 \label{5},\ ] ] where @xmath12 are the field dependent masses of the stop particles and @xmath13 is the field dependent top mass .
the scale @xmath4 remains arbitrary and is usually chosen to be equal to the value of the top mass .
in fact the potential is scale independent since explicit dependence on the scale is compensated by the implicit dependence of the parameters renormalized at this scale .
@xmath14 is the value of the potential at @xmath15 , which has to be subtracted in order to keep the scale invariance @xcite .
the field dependent squarks masses are given by the eigenvalues of the mass - squared matrix @xmath16 where @xmath17 is the conventional trilinear soft supersymmetry breaking parameter , @xmath18 is the higgs mixing parameter , and the top mass squared is given by @xmath19 .
the so - called d - terms give contribution proportional to the gauge couplings and will be ignored hereafter in order to gain approximate scale independence of the potential , since we are including only top - stop contributions to @xmath20 .
then the eigenvalues of the matrix ( [ 6 ] ) are @xmath21 , \ ] ] where @xmath22 .
the scale independence of the effective potential ( [ 5 ] ) is given by the renormalization group equation @xmath23 where @xmath24 are the parameters of the tree - level potential and @xmath25 s denote their @xmath26 functions .
according to the approximation mentioned above , we have ignored the scale dependence of the gauge couplings .
the general solution of eq.(7 ) has the form : @xmath27 where the function @xmath28 is the perturbative expression with the scale chosen arbitrary , @xmath24 are the effective parameters of the potential , @xmath29 and @xmath30 s are anomalous dimensions of the fields @xmath31 .
in particular one can choose @xmath4 to be equal to the top mass , as is usually done .
however , the main distinction is whether we choose the top mass to be field dependent and put it equal to its numerical value after the minimization of the potential , or we take its numerical value at the very beginning and then minimize the potential . in the latter case
we incorporate the perturbative corrections to the potential , while in the first case we sum all the leading logs via the renormalization group equation . proceeding the first way we get @xmath32 \nonumber,\end{aligned}\ ] ] where @xmath13 is the field dependent mass and @xmath33 and @xmath34 are the stop masses boundary values when @xmath15 .
eq.[9 ] is the rg improved expression for the one - loop effective potential which corresponds to taking into account all the leading log contributions proportional to the top yukawa coupling from all the loops .
to find the vacuum expectation values of the higgs fields we have to minimize the potential . using the notation @xmath35 and keeping only the terms of the first order in coupling constants the minimum of the potential ( [ 9 ] )
is given by @xmath36\tan^2\beta+(a_t^2\tan^2\beta -\mu^2 ) \frac{f(\tilde{m}^2_{t1})-f(\tilde{m}^2_{t2})}{\tilde{m}^2_{t1}- \tilde{m}^2_{t2}}\right]\bigg\ } , \label{10 } \\
2m_3 ^ 2&=&\frac{\sin 2\beta}{1+\varepsilon_1}\bigg\{m_1 ^ 2(1-\gamma_2 ) + m_2 ^ 2(1-\gamma_1)-\varepsilon_2 \label{11 } \\ & + & \frac{3h_t^2}{16\pi^2}\left[f(\tilde{m}^2_{t1 } ) + f(\tilde{m}^2_{t2})-2f(m^2_t)+(a_t + \mu\tan\beta)(a_t + \mu\cot\beta ) \frac{f(\tilde{m}^2_{t1})-f(\tilde{m}^2_{t2})}{\tilde{m}^2_{t1}- \tilde{m}^2_{t2}}\right]\bigg\ } , \nonumber\end{aligned}\ ] ] where @xmath37 , \\ \gamma_1&=&-\frac{3}{2}(\tilde{\alpha}_2 + \frac{1}{5}\tilde{\alpha}_1 ) , \\
\gamma_2&=&\frac{3}{2}(2y_t-\tilde{\alpha}_2 -\frac{1}{5}\tilde{\alpha}_1 ) , \\ y_t&=&\frac{h_t^2}{16\pi^2 } , \tilde{\alpha}_i=\frac{\alpha_i}{4\pi}=\frac{g_i^2}{16\pi^2}.\end{aligned}\ ] ] here the values of all the mass parameters and couplings are taken at the scale equal to the top mass .
having in mind eqs.([11 ] ) we are now in a position to calculate the rg improved radiative corrections to the masses .
one has : @xmath38 where @xmath39 and @xmath40 is the one - loop radiative correction @xcite , @xcite @xmath41 .\ ] ] using eqs.([12 ] ) we can also calculate the corrections to the squared masses in the cp - odd neutral sector . just like in the usual case @xcite taking the second derivative of the full potential with respect to @xmath42
one has : @xmath43 where @xmath44 this gives for the cp - odd higgs mass @xmath45 where @xmath46 and @xmath47\ ] ] for the cp - even sector we have to differentiate the potential with respect to @xmath48 we find @xmath49 @xmath50 where @xmath51 , \\ \delta_6&=&6
\left[(\tilde{\alpha}_2m_2 ^ 2+\frac{1}{5}\tilde{\alpha}_1m_1 ^ 2 + \tilde{\alpha}_2\mu^2+\frac{1}{5}\tilde{\alpha}_1\mu^2)(\cot^2\beta-1 ) + 2(\tilde{\alpha}_2m_2+\frac{1}{5}\tilde{\alpha}_1m_1 ) \mu\cot\beta \right .
\\ & & \left .
-am_0\mu y \cot\beta \right]\end{aligned}\ ] ] and @xmath52 s are @xcite : @xmath53 ^ 2d(\tilde{m}^2_{t1},\tilde{m}^2_{t2 } ) , \\
\delta_{22}&=&\frac{3g^2}{16\pi^2}\frac{m^4_t}{\sin^2\beta m^2_w}\left [ \ln ( \frac{\tilde{m}^2_{t1}\tilde{m}^2_{t2}}{m^4_t})+ \frac{2a_t m_0 ( a_t m_0 + \mu\cot\beta ) } { \tilde{m}^2_{t1}-\tilde{m}^2_{t2 } } \ln ( \frac{\tilde{m}^2_{t1}}{\tilde{m}^2_{t2 } } ) \right.\\ & & \left.+\left [ \frac{a_t m_0(a_t m_0 + \mu\cot\beta ) } { \tilde{m}^2_{t1}-\tilde{m}^2_{t2}}\right]^2 d(\tilde{m}^2_{t1},\tilde{m}^2_{t2})\right ] , \\
\delta_{12}&=&\frac{3g^2}{16\pi^2}\frac{m^4_t}{\sin^2\beta m^2_w } \frac{\mu(a_t m_0 + \mu\cot\beta ) } { \tilde{m}^2_{t1}-\tilde{m}^2_{t2 } } \left [ \ln ( \frac{\tilde{m}^2_{t1}}{\tilde{m}^2_{t2 } } ) + \frac{a_t m_0 ( a_t m_0+\mu\cot\beta ) } { \tilde{m}^2_{t1}-\tilde{m}^2_{t2 } } d(\tilde{m}^2_{t1},\tilde{m}^2_{t2})\right ] \\\end{aligned}\ ] ] @xmath54 @xmath55 is the mass of a light squark .
the diagonalization of the matrix ( [ a ] ) gives us the masses of the cp - even neutral higgses .
for the charged ones one has the usual expression @xcite @xmath56 where @xmath57 and the only difference is that @xmath58 is given by eq.([15 ] ) .
resulting expressions for the higgs masses differ from those obtained without rg summation . to calculate the corrections one has to perform the usual procedure @xcite , @xcite , @xcite of fitting the set of soft breaking parameters , @xmath59 as well as the top mass @xmath13 .
this is not a straightforward operation , since one has to fulfill many requirements simultaneously , thus defining the optimized best fit @xcite .
having performed the complete procedure with the help of computer programm we have got the best fit values of the parameters mentioned above and used them in our formufae .
one of the important observation is the character of @xmath13 dependence of the results . comparing with that obtained without account of the rg summation @xcite
, we find it to be smoother .
fig.1 shows our results .
we conclude that the account of rg summation procedure can introduce the changes in the predictions of the higgs masses .
one can observe from fig.1 that there exist the lower bound on the higgs mass of about 95 gev . in the interval of top mass preferable according to the recent cdf @xcite data ( @xmath61 ) , that corresponds to the running top mass of 164 @xmath62 , the @xmath13 dependence of the lightest higgs mass is very weak and @xmath63 appears to be lighter than 100 gev .
the numerical analysis was performed with the help of the computer program , developed in @xcite .
we are grateful to w.de boer and r.ehret from karlshrue university for valuable discussions and for necessary modification of the program . | the one - loop radiative corrections to the higgs boson potential in the mssm , originating from the top quark and squark loops , are summed in the leading log approximation using the renormalization group .
the rg improved effective potential is minimized and the corrections to the cp - odd and cp - even higgs boson masses are calculated .
the resulting masses exhibit smoother top mass dependence than those calculated without rg summation . we have also found that for preferable values of the top mass the light higgs mass does not exceed 100 gev . 6.3 in .2 in .2 in jinr - e2 - 94 - 400
+ october 1994 * renormalization group improved radiative corrections + to the supersymmetric higgs boson masses * |
this material consists of three parts . in the first part
we show explicitly how to find stationary values @xmath79 of the ssrs from eq . using a simple toy two - level model of the qd . in the second part
, we provide a detailed example of the averaged quantities , lasing intensity and current , in dependence on voltage as promised in the main text .
finally , we briefly consider a device made by placing many josephson leds into the same optical cavity .
a presentational problem with the setup described in the main text , is a relatively large number of qd states even when using all conservation laws ( @xmath132 states for the odd parity or @xmath133 states for the even parity ) . to circumvent this problem and
still discuss the essential properties of the self - consistency equation and the novel driving mechanism , let us consider a toy two - state model .
the great advantage of this approach is that all calculations can be performed explicitly . _
two - level system _ for the toy model , we take one of the states to be @xmath134 , which is unaffected by the superconductivity in the leads .
the other state is chosen as a superposition of two states with an even number of particles in each dot , @xmath135 .
these number states are mixed by the sc leads with an angle @xmath136 .
the phase @xmath54 is the sc phase difference .
the hamiltonian @xmath137 in this subspace is of the form @xmath138 here , @xmath139 is the energy difference between the two states in the absence of radiation and @xmath140 is due to the dipole coupling to the resonant mode .
diagonalizing this hamiltonian yields the eigenenergies @xmath141},\ ] ] where ` @xmath142 ' labels the two eigenstates of eq . .
note that for @xmath76 we have @xmath143 and @xmath144}}.\ ] ] the dipole serves as a driving force in the self - consistency equation which for stationary states is given by @xmath145 as was noted in the main text , a laser field can develop when the energy gain rate @xmath146 $ ] due to the josephson led is larger than the energy loss rate @xmath147 . with increasing @xmath48 the energy gain will saturate because the denominator of eq .
increases , until a _ stationary state of radiation _ ( ssr ) , @xmath53 , is reached ( @xmath148 ) and steady state lasing is achieved . for this toy model
the energy gain is given by @xmath149}}\end{aligned}\ ] ] where @xmath150 is the optical phase , @xmath151 .
_ driving mechanism _ in the main text we have stressed the novelty of the hjl driving mechanism . in conventional lasers ,
the driving is due to a population inversion that originates from dissipative transitions in an open system .
however , for the hjl the drive not dissipative and the energy gain depends crucially on mixing of qd states by the induced sc gaps without any population inversion . in our toy - model
the essence of the driving comes about through the mixed state @xmath152 which renders both @xmath153 and @xmath154 non - zero .
therefore , irrespective of the eigenstate of the qd , the system can always decay to the other state and emit a photon . without mixing ( @xmath155 ) this is not possible .
the energy required for the decay is supplied by the bias . with increasing @xmath48 the values of the amplitudes decrease as the eigenstates
@xmath156 will become increasingly more like the eigenstates of @xmath42 . hence the driving saturates . from the expression for the energy gain , we explicitly see the role of the sc phase difference , @xmath54 .
first , the value of this phase determines the sign of the energy gain , and thus whether it acts as a gain or as an absorber of radiation .
second , the ssrs with radiation amplitudes , @xmath53 , depend on the specific combination of phases , @xmath157 .
hence , the ssr values of the optical phase must depend on the sc phase difference , which signifies a phase - lock between the optical phase and sc phase difference .
we note that , because of this specific combination of phases , there are always two optical phases possible since changing @xmath158 yields an ssr with the same radiation magnitude @xmath56 but opposite sign . hence , the ssrs come in pairs @xmath55 .
this is in stark contrast to conventional lasers where the driving is independent of the optical phase and all phases occur with equal probability . as a final remark concerning the drive we note that the dipole @xmath159 completely saturates in the limit of @xmath160 , to @xmath161 in this limit the eigenstates of @xmath162 are the eigenstates of @xmath42 so that the amplitudes @xmath163 and @xmath164 , we need to solve the stationary self - consistency equation with the dipole of eq . .
here we see by inspection that @xmath76 is always a stationary solution since the dipole is proportional to @xmath165 .
it may be unstable however against small fluctuations in the mode . by performing stability analysis
we distinguish between the two cases .
to find a ssr with nonvanishing radiation amplitude , we first divide both sides of eq .
( taking dipole @xmath166 ) with its conjugate to derive expressions for the optical phase @xmath167 where @xmath168 .
this reduces to @xmath169 with extra condition @xmath170 = 1 $ ] . from this
, we explicitly see that the optical phase @xmath150 may assume two equivalent values shifted by @xmath171 and that it is locked to sc phase @xmath54 .
nontrivial solutions for the optical phase @xmath150 exist for @xmath172 , i.e. , only at negative detuning . for the ` @xmath173 ' eigenvalue
, the situation is reversed in detuning : the lasing solutions exist only at positive @xmath39 with @xmath174 . now taking modulus square of both sides of
, we arrive at an expression for the optical field magnitude @xmath175 } { 1+\sin^2(2\theta ) + 2 \sin(2\theta)\cos(2\chi-\phi))}.\ ] ] solving eqs .
and simultaneously yields all ssrs , @xmath176 . from eq .
we may derive expressions for the lasing thresholds as it implies that the lasing solutions with @xmath177 can only exist above some critical coupling @xmath178.\ ] ] stability analysis shows that the zero stationary solution becomes unstable at this critical coupling . because the lasing field continuously starts to grow when crossing this threshold , we have the analog of a second - order phase transition .
interestingly , at yet higher coupling @xmath179\ ] ] the zero solution becomes stable again and we encounter an analog of the first - order transition where the ssrs with nonvanishing radiation amplitudes still exist at the threshold . at the tricritical point @xmath180 example of current / intensity dependence on voltage -------------------------------------------------- , in the main text .
the solid curves represent the time - averaged currents which are the individual currents weighted with the probabilities @xmath102 , the stationary solutions to master equation of the switching dynamics as explained in the text .
the qd parameters are as given in the main text , @xmath181 and @xmath69 , and @xmath182 . ] in the main text the dynamics of the hjl over longer times was described in terms of switching rates .
it was argued that these switchings occur after such long times that the corresponding dynamics is decoupled from the dynamics of the lasing field @xmath48 .
hence , the switchings always occur when the system is at an ssr of eq . , @xmath79 , where the qd is in eigenstate @xmath46 .
we additionally argued that ( for @xmath183 ) switchings that change the parity of the qd state occur at an even slower rate . even for @xmath184
slower than the switchings @xmath85 . therefore ,
when measuring over times much longer than @xmath185 but shorter than @xmath186 , we see an average intensity corresponding to the ssrs of many qd eigenstates with equal parity . after a typical time scale @xmath186
the qd states switch parity and the average intensity corresponding to this parity is seen .
the purpose of this section is to calculate the average intensity as a function of voltage for the even and odd parities separately and discuss some important features found in the corresponding curves .
the average intensity can be found by solving the master equation governing the switching dynamics and finding the corresponding stationary solution .
the probability @xmath187 to be at @xmath79 evolves in time according to @xmath188\equiv\sum_{n , r}\mathbf{w}_{m , s ; n , r}\ p_r^n\ ] ] where @xmath189 denotes the transition from @xmath190 and we define the transition matrix @xmath191 .
the summation is over the stationary solutions @xmath192 of all eigenstates , @xmath193 . to solve master equation we need to find and diagonalize @xmath194 .
we notice that this matrix has a left eigenvector with zero eigenvalue because @xmath195 .
hence there must also be a right eigenvector with zero eigenvalue , which is the stationary distribution of probabilities ( @xmath196 for all @xmath72 ) when normalized to unity .
since we are only interested in these stationary solutions we only need to find the right null vector of @xmath191 .
we represent this stationary solution as @xmath102 , with a capital @xmath197 .
hence , we can find @xmath102 when all @xmath198 are known . to calculate the transitions
@xmath189 we note that the switchings are accompanied by the emission of an off - resonant photon so that the rates follow from fermi s golden rule .
the transition rates are then proportional to the product of @xmath199 $ ] , with @xmath200 the energy difference between @xmath201 and @xmath202 .
we note that the transitions go from eigenstate @xmath203 ( @xmath204 ) , which are the eigenstates of the hamiltonian @xmath205 in the semiclassical approximation , @xmath206 .
additionally we stress that the value of @xmath48 is unaltered during the transition , which is indicated by the @xmath207 in @xmath208 and @xmath201 .
only after the switch will @xmath48 evolve to @xmath79 according to eq . .
to find the matrix elements of @xmath209 [ eq .
] we only need to consider the interactions @xmath210 as the term @xmath211 will always yield zero contribution when @xmath204 .
the two interactions imply two possible transition amplitudes distinguishing two types of emission : one far below and one far above the cavity resonance .
since we are in the rotating frame , this translates to creating a negative frequency photon with @xmath212 , or creating a positive frequency photon with @xmath213 .
we do not expect absorption of off - resonant photons as they are very scarce .
taking the detuning into account , the photon energies are @xmath214 and @xmath215 respectively .
the detuning may be neglected , however , as it is typically taken of the same order as @xmath216 .
for this same reason we may also neglect the @xmath4 in the denominator of the density of states @xmath217 . taking the above considerations into account we find the transition rates @xmath218 with this we can find @xmath219 and thus @xmath102 .
we make some brief remarks about @xmath219 .
as stipulated above the switchings only occur between different qd eigenstates .
fluctuations that let the system jump from one to another ssr in the same qd eigenstate are exponentially small due to the large number of photons .
thus @xmath220 if @xmath221 .
furthermore , we note that when in a lasing eigenstate and starting at @xmath76 the system may evolve to both ssrs with @xmath222 with equal probability .
hence the transition rate from a non - lasing qd eigenstate to the ssr with @xmath79 of a lasing qd eigenstate is half the value calculated in eq . .
for transitions between two lasing eigenstates it must be determined to which of the multiple stationary points @xmath79 the field evolves .
this is done by integrating the time evolution equation of @xmath48 , eq . , until one of the @xmath79 is sufficiently approached , using the radiation amplitude @xmath223 of the previous ssr as initial condition . with the stationary probability distribution , @xmath102
, we calculate some measurable quantities .
first , the average number of photons in the cavity is @xmath104 .
then the average number of photons that escape the cavity is @xmath224 , yielding the average emission power @xmath225 and the average current in the device @xmath226 .
the latter because every photon emission is accompanied by a single charge transfer . using the minimal model of the main text
we calculate the current / intensity as a function of detuning .
as mentioned above we plot the dependencies for odd and even parity states separately .
the results are presented in fig .
[ fig : curves ] .
since the detuning is related to the applied voltage via @xmath227 , the peaks in the current , of the order of @xmath107 , are in a narrow interval @xmath108 of voltages .
there are two kinds of discontinuities in the curves : kinks , that mark the thresholds of the lasing instability at @xmath76 in some qd eigenstate ( second - order transitions ) , and jumps , that signal the appearance of an ssr far from @xmath76 ( first - order transitions ) .
the curves also show a relatively high probability to be in lasing qd eigenstates if available .
this can be qualitatively understood by considering that eigenstates at large @xmath48 are close to eigenstates of @xmath110 .
this suppresses off - diagonal matrix elements , eq . , resulting in a slower rate of transitions away from a lasing state . for the odd parity states in fig .
[ fig : curves ] , the time - averaged current drops to zero while a lasing state is still available .
this is because the non - lasing solution in this lasing eigenstate becomes stable at @xmath228 and whereas all other states are nonlasing .
the device then gets stuck in the ssrs at @xmath76 even though ssrs with @xmath229 are available .
we have discussed in the main text that the large number of photons @xmath230 ( assuming @xmath231 ) in a cavity with a single josephson led requires high @xmath122 factors .
this requirement can be easily softened , and the lasing can be achieved at higher damping rate @xmath4 , if we incorporate @xmath232 josephson leds in the cavity .
since the leds are of ten nanometer scale , one can put hundreds of them even into a single - wavelength cavity .
the resulting system certainly deserves to be explored in detail , and will be a topic of separate research presented elsewhere . in this note
we just give some straightforward estimations .
all @xmath233 qd dipoles are in this setup coupled to a single mode , were we assume equal coupling coefficients @xmath62 .
the self - consistency equation for the radiation amplitude changes to @xmath234 assuming for simplicity all dipoles to be in the same direction with same or similar parameter values , we may simplify the sum over dipoles to @xmath235 .
this yields @xmath236 for the ssrs .
the optimal regime near the lasing threshold is achieved if @xmath237 .
thus this corresponds to @xmath238 .
therefore , the lasing threshold can be achieved at much larger damping rate @xmath239 .
this also increases the number of photons in the mode .
if we optimize the coupling to @xmath238 , the number is estimated as @xmath240 , @xmath233 times larger than the estimate for a single - led device . if we fix @xmath4 , @xmath62 and increase the number of wires in the device , the photon number scales as @xmath241 .
note that in both cases the emitted power increases as well .
it is respectively given by @xmath242 , for optimal coupling , and @xmath243 when only increasing the number of wires .
it is likely that a large number of leds in the cavity greatly reduces the intensity fluctuations . with only one pair of qds
the lasing intensity fluctuates down to zero as can be seen from fig .
[ fig : timeline ] , in the main text .
in contrast , the switching of only one of @xmath233 qds will hardly affect the intensity , so the relative intensity fluctuations are expected to be of the order of @xmath244 .
likewise , the phase of the optical field will be barely affected by the switching in one nanowire .
therefore the decoherence times may become significantly longer than in a single led hjl . | we describe a superconducting device capable of producing laser light in the visible range at half of the josephson generation frequency with the optical phase of the light locked to the superconducting phase difference .
it consists of two single - level quantum dots embedded in a _
p - n _ semiconducting heterostructure and surrounded by a cavity supporting a resonant optical mode .
we study decoherence and spontaneous switching in the device .
lasers and superconductors are both systems with macroscopic quantum coherence . in lasers ,
photons form a coherent state induced by stimulated emission of a driven system into a cavity mode .
the resulting visible coherent light is characterized by an optical phase @xcite . in superconductors ,
the ground state arising from spontaneous symmetry breaking is also characterized by a phase @xcite .
traditionally , lasers and superconductors are studied separately .
recently @xcite , it has been realized that the superconducting ( sc ) phase difference and the optical phase may interact in a single device that combines two superconductors and a semiconducting _ p - n _ junction .
the latter is a common system for light generation as the electron - hole recombination produces photons of visible frequency @xcite . combining semi- and superconductors within a nanostructure
has been a difficult technological problem that attracted attention for a long time @xcite .
it has been solved using semiconductor nanowires @xcite or quantum wells @xcite , opening up the possibility to make combined devices .
the device in question has been termed a josephson led [ fig .
[ fig : setup](a ) ] .
it employs a double quantum dot ( qd ) in a _
p - n _ semiconductor nanowire connected to sc leads @xcite .
the device , biased with a voltage @xmath0 , exhibits two types of photon emission : `` blue '' photons at the josephson frequency @xmath1 due to the recombination of a cooper pair from each side of the junction , and `` red '' photons at about @xmath2 due to electron - hole recombination .
it has been shown that the optical phase of the josephson generated `` blue '' photons is locked with the sc phase difference .
the resulting `` blue '' light could in principle be enhanced by traditional optical methods but its small intensity makes this a challenging task . in this letter , we explore an alternative idea where the far more intense `` red '' emission is enhanced in a resonant cavity mode .
we find lasing at half the josephson frequency and , thus , dub the device ` half - josephson laser ' ( hjl ) . in a common laser , lasing results from spontaneous symmetry breaking where all values of the optical phase are equivalent .
drift between these values leads to a finite decoherence time .
in contrast , the optical phase of the hjl is locked to the sc phase difference with only two allowed values of the optical phase corresponding to two opposite radiation amplitudes .
this removes drift as a source of decoherence and opens up the possibility to manipulate the optical phase by changing the sc phase difference . instead , decoherence of the radiation in the hjl results from switching between different qd states accompanied by the emission of a photon .
we have explored these processes and find that by order of magnitude the resulting decoherence time is the same as the theoretical limit for a common laser @xmath3 , with @xmath4 the damping rate and @xmath5 the number of photons accumulated in the resonant mode .
a rather low @xmath4 is required to achieve lasing for a single josephson led , this condition being relaxed with a large number of leds in a single cavity @xcite . of the sc leads which differ by an energy @xmath6 .
charge transfer is only possible either through electron - hole recombination with the emission of a `` red '' photon at @xmath7 or through a cooper pair transfer with the emission of a `` blue '' photon at @xmath8 .
( b ) the hjl is a josephson led embedded in an optical cavity with a resonance frequency @xmath9 , i.e. , close to the `` red '' emission frequency .
the separately - colored regions in between the depleted areas represent the two qds . ]
_ setup and model _ the hjl is a josephson led embedded in a single mode optical cavity with resonance frequency @xmath10 , fig .
[ fig : setup](b ) . the light emission from the cavity
is described by a damping rate @xmath4 .
the electronic part consists of a biased _ p - n _ junction where each side of the junction accommodates a qd connected to a sc lead .
the barriers separating the qds from the leads are arranged such as to allow charge transfer only through electron - hole recombination .
such qd junctions can be realized with semiconducting nanowires @xcite .
the minimal model for the qds involves a single orbital for each qd . an orbital can house up to two particles ( including spin ) yielding 16 possible states .
the qd hamiltonian then reads @xcite @xmath11 + u_{eh } \hat n_e \hat n_h,\ ] ] where @xmath12 ( @xmath13 ) is the electron ( hole ) number operator , and @xmath14 ( @xmath15 ) is the annihilation operator for an electron ( hole ) with spin @xmath16 .
the energies @xmath17 are measured with respect to chemical potentials @xmath18 of the corresponding leads that differ by an energy @xmath19 ; here , @xmath20 is the on - site charging energy and @xmath21 the coulomb attraction between electrons and holes . for concreteness , we assume that the hole level houses a heavy hole with @xmath22 , where @xmath23 is the nanowire axis @xcite .
such levels are commonly used in optical experiments with qds @xcite .
our qualitative results do not depend on this particular choice .
due to the proximity of the sc leads , cooper pairs can coherently tunnel between the sc leads and the qds introducing mixing between unoccupied and doubly occupied qd states .
these processes can be compactly described by an additional term @xmath24 in the hamiltonian ; here , the induced pair potentials @xmath25 have reduced magnitudes in comparison with the gaps @xmath26 of the sc leads but they retain the same phases @xmath27 . owing to gauge invariance
, the physical quantities depend only on sc phase difference @xmath28 the hamiltonian is valid under the conditions @xmath29 .
we note further that this hamiltonian along with the electron - hole recombination conserves parity ( _ even _ or _ odd _ ) of the total number of particles on the qds .
even - odd transitions require creation of quasiparticle excitations in the sc leads and occur with a relatively slow rate estimated below .
interaction between the resonant mode and qds is described by @xmath30 ; @xmath31 being electric field of the mode at qd position and @xmath32 the dipole moment of the optical transition between the conduction and the valence band .
we assume a linear polarized mode , choose the @xmath33-axis in the direction of the polarization , and notice that for heavy holes @xmath34 with @xmath35 .
the time - dependence of the dipole moment is due to the applied voltage .
it is convenient to implement a rotating - wave approximation transferring the time - dependent factor to the photon creation ( annihilation ) operator @xmath36 ( @xmath37 ) .
thereby , the photon - dependent part of the hamiltonian reads @xmath38 with @xmath39 being the frequency _ detuning _ , @xmath40 , @xmath41 .
we see that @xmath42 plays the role of a driving force that excites the oscillations in the mode .
we note that all hamiltonians considered conserve spin .
_ semiclassics _ the present model is a rather complex case of nonequilibrium dissipative quantum mechanics .
however , since we envisage a large number of photons in the mode , we employ a semiclassical approximation replacing @xmath43 .
the hamiltonian @xmath44 can then be diagonalized to obtain the spectrum @xmath45 and corresponding eigenstates @xmath46 . the dipole strength @xmath47 depends both on the radiation field @xmath48 and the qd state @xmath46 .
since the dipole strength in turn determines the evolution of the radiation field via the evolution equation @xmath49 we have to solve the system self - consistently @xcite .
the radiation field can build up as long as the energy gain rate @xmath50 $ ] due to the nanowire is greater than the energy loss rate @xmath51 . with increasing @xmath48 the energy gain saturates till a _ stationary state of radiation _ ( ssr ) with @xmath52 is reached at a certain radiation amplitude @xmath53 . in conventional lasers , the driving is due to a population inversion that originates from dissipative transitions in an open system . for the hjl ,
the sc drive is not dissipative by itself : only the emission of photons from the cavity is a dissipative process .
the driving originates from coherent mixing of discrete quantum states due to the proximity of the qd to the sc leads without any population inversion .
thus , the driving mechanism of the hjl is very different from that of a conventional laser .
this is why the information about the sc phase difference is preserved in the process of driving .
the energy gain , including its sign , depends on the difference between @xmath54 and the phase of @xmath48 .
owing to this , the phase of @xmath53 of the ssr is locked to the sc phase difference .
the ssrs of the hjl come in pairs @xmath55 which is very different from a conventional laser where only the magnitude @xmath56 ( photon number ) is fixed .
we give in @xcite analytical solutions to eq . for a toy two - level model .
_ scales
_ let us estimate the scales involved that are expected to yield lasing . to simplify , we assume all characteristics of the qd spectrum to be of the same energy scale @xmath57 which is of the order @xmath58 .
this assures optimal mixing of the qd states by superconductivity . in a lasing state
, the radiation amplitude should noticeably contribute to the energies of the qd states .
this requires @xmath59 . assuming @xmath60 we estimate from eq . that this takes place at @xmath61 .
we will assume that @xmath62 is always chosen to be of this scale .
the number of photons is then estimated as latexmath:[$n \simeq is thus justified provided @xmath4 is sufficiently small , @xmath64 .
_ lasing _ despite the model being minimal , it contains ten parameters that affect the existence and characteristics of the ssrs . to find these characteristics , we need to evaluate the dipole moment in a given state at given @xmath48 , and can do it separately for the states of _ odd _ and _ even _ parity since they are not mixed by interactions
additionally , the spin conservation splits the eight odd states into two equivalent groups of four corresponding to total spin @xmath65 . for the even states , only one of the four possible @xmath66 states , @xmath67 , couples to the field .
hence , we only need to consider five of the eight even states as the other three are dark . with this , we demonstrate lasing as proof of concept by finding ssrs in the even states for qd parameters within the above estimated scales , @xmath68 and @xmath69 , for wide regions in the space of detuning @xmath39 and coupling @xmath62 , see fig . [
fig : stability ] .
note that each eigenstate @xmath46 has a different dipole strength @xmath70 such that the lasing threshold @xmath71 [ fig .
[ fig : stability](a ) ] and the radiation amplitudes of the ssrs @xmath53 [ fig .
[ fig : stability](c ) and ( d ) ] depend on @xmath72 .
figure [ fig : stability](b ) shows the number of photons upon crossing the lasing threshold for the state @xmath73 . in agreement with the estimations
, @xmath5 reaches the maximum @xmath74 at @xmath75 .
we stress that the optical phase of the radiation amplitude in an ssr is not arbitrary but locked to the sc phase difference [ fig .
[ fig : stability](c ) ] . for three eigenstates
at the line with the number @xmath72 the energy gain at @xmath76 for the state @xmath46 exactly equals the energy loss .
( b ) number of photons @xmath5 for eigenstate @xmath73 versus the coupling constant @xmath62 for @xmath77 [ dotted line in ( a ) ] above the lasing threshold at @xmath78 .
plots ( c ) and ( d ) illustrate the radiation amplitudes @xmath79 ( marked with circles ) for ssrs corresponding to the different eigenstates of the qd .
the parameter choice is given by the cross in ( a ) where only the states @xmath73 and @xmath80 are lasing .
( c ) non - zero @xmath79 come in pairs with opposite sign .
changing the sc phase difference will rotate the @xmath79 with respect to the origin of the plot .
( d ) eigenenergies versus @xmath48 ( at @xmath81 ) .
the @xmath79 are different for each @xmath46 . ]
_ switching _ in the above discussion , we have assumed the qd to stay in a certain eigenstate @xmath46 .
in fact it does not : the finite spectral width of the mode enables switching between the eigenstates .
as shown below , the switching events occur on a much longer timescale @xmath82 than that of the relaxation of @xmath48 towards its stationary value , @xmath83 .
this separation of timescales allows to consider the switching dynamics separately from the dynamics of the radiation amplitude .
each switching event is accompanied by the emission of a photon with a frequency mismatch compensating the difference of energies between initial and final eigenstates , @xmath84 . for switchings not altering the parity of the eigenstate , the rates @xmath85 can be evaluated using fermi s golden rule that contains the effective density of photon states @xmath86 ( the tail of a lorentzian - shaped emission line of the resonant mode ) and the square of the matrix element , @xmath87 , with @xmath88 denoting the initial ( final ) state .
thereby , the switching rate can be estimated as @xmath89 .
the switching events are thus rare and the device stays in one of the ssrs between the events .
switchings altering parity are even rarer as they require the excitation of a quasi - particle above the sc energy gap @xmath90 .
the larger detuning of the off - resonant photon @xmath91 and an additional small factor @xmath92 result in a parametrically smaller rate @xmath93 @xcite .
such processes do not conserve spin thus enabling switchings between dark and emitting states .
it is important to realize that , since the ssrs for different eigenstates have different values of @xmath79 , @xmath48 does not jump to the new stationary value upon a switching .
rather , the amplitude will evolve to @xmath94 within a timescale @xmath95 , according to eq . .
for the same reason a switching event always involves different eigenstates rather than different ssrs at the same eigenstate .
the latter would require large fluctuations of @xmath48 that are suppressed exponentially .
fig .
[ fig : timeline ] shows a sketch of the radiation intensity as function of time .
in contrast to common lasers , the hjl intensity fluctuations are large at timescales of @xmath82 .
_ decoherence _ the intrinsic mechanism of decoherence in common lasers is a drift of the optical phase .
for the hjl , this mechanism does not work since the amplitudes of the ssrs are locked to the sc phase difference .
this renders switching the most important source of decoherence in the hjl .
indeed , after switching from a lasing to a non - lasing ssr the radiation extinguishes quickly and its phase is forgotten .
even if the next switching brings the system to a lasing eigenstate , the radiation will evolve from the initial @xmath76 to any of the two possible @xmath96 , with equal probability .
since decoherence is due to switching , the relevant timescale is given by @xmath97 . despite the very different decoherence mechanism ,
this estimation is the same as for the common laser @xcite . evolving in time .
( a ) switching events not altering the parity of the qds occur at the timescale @xmath98 . after switching
, the radiation amplitude attains its new stationary value at a timescale @xmath99 during which the qd remains in the same eigenstate .
( b ) switching events altering the parity occur at a longer timescale @xmath100 .
these can change between the dark and lasing states . ] _ average power and current _ the intensity fluctuations due to switching self - average at timescale exceeding @xmath101 .
the averaged characteristics are expressed in terms of the probabilities @xmath102 to be in a ssr @xmath103 that belongs to an eigenstate @xmath46 . those are given by the stationary solution to the master equation of the switching dynamics , that is composed of the switching rates @xcite . in terms of these probabilities , the average number of photons in the cavity is given by @xmath104 .
the average emission power is proportional to the photon number , @xmath105 .
the same holds for current in the device : since the emission of each photon is accompanied by a charge transfer , it is given by @xmath106 .
an elaborated example of the current / intensity dependences is provided in @xcite . for the current - voltage characteristic
, we find a rather complex structure beside a peak with a magnitude of the order @xmath107 , that is concentrated in a narrow interval @xmath108 of voltages in the vicinity @xmath109 .
in this structure , two types of discontinuities are present : ( _ i _ ) kinks marking the thresholds of lasing instability at @xmath76 ( second - order transitions ) and ( _ ii _ ) jumps signaling the appearance of a ssr with stationary radiation amplitude @xmath53 far from @xmath76 ( first - order transitions ) .
we observe a relatively high probability to remain in an ssr with a large photon number .
it is explained from the fact that eigenstates at large @xmath48 are close to eigenstates of @xmath110 .
this suppresses off - diagonal dipole - matrix elements resulting in a suppressed rate of transitions from this state . _
feasibility _ to show the feasibility of the hjl , we present here the estimations with concrete numbers .
the sc gaps @xmath90 are typically @xmath111mev , so we can choose the qd energy scale @xmath112mev . to estimate the dipole strength @xmath113 , with @xmath114 being the quantum fluctuation of the electric field in the mode , we assume the cavity volume @xmath115 with the wavelength @xmath116 , and take @xmath117 for the atomic distance scale .
this gives the maximum @xmath118mev . with these two values for @xmath57 and @xmath62 ,
the minimum damping rate required for lasing is @xmath119hz , corresponding to quality factor @xmath120 which is common for optical cavities .
however , in this situation the number of photons @xmath121 . this can be enhanced by increasing @xmath122 and simultaneous decreasing @xmath62 so it remains @xmath123 .
for photonic crystal cavities @xcite quality factors @xmath124 @xcite have been measured and @xmath125 @xcite have been theoretically predicted .
this gives photon numbers @xmath126 and @xmath127 , respectively .
the estimations of the emitted power and current at the peak do not depend on the choice of @xmath4 and are given by @xmath128nw , @xmath129na .
the requirements on @xmath4 can be eased and @xmath5 enhanced by putting many josephson leds in the same cavity .
furthermore , this also increases the emission power @xmath130 and the current @xmath131 @xcite . in conclusion , we have demonstrated the feasibility of generating coherent visible light at half the josephson frequency in a sc nanodevice .
the workings of the device resemble the spontaneous parametric down - conversion in nonlinear optics @xcite with the superconductors playing the role of coherent optical input . the novel driving mechanism results in locking between optical phase and sc phase difference .
the decoherence of the emitted light originates from the switchings between different quantum states of the device .
we acknowledge fruitful discussions with n. akopian , l.p .
kouwenhoven , m. reimer , and t. van der sar and financial support from the dutch science foundation nwo / fom . 10 m.o .
scully and m.s .
zubairy , _ quantum optics _ ( cambridge university press , cambridge , 1997 ) .
m. tinkham , _ introduction to superconductivity _ ,
2nd edition ( mcgraw - hill , new york , 1996 ) .
p. recher , yu.v . nazarov and l.p .
kouwenhoven , phys .
rev .
lett .
* 104 * , 156802 ( 2010 ) .
c. weisbuch and b. vinter , _ quantum semiconductor structures : fundamentals and applications _
( academic press , san diego , 1991 ) .
h. takayanagi , t. akazaki and j. nitta , phys .
rev .
lett . * 75 * , 3533 ( 1995 ) ; s. de franceschi _ et al .
_ , appl .
phys . lett . *
73 * , 3890 ( 1998 ) .
y.j .
doh _
et al . _ ,
science * 309 * , 272 ( 2005 ) ; j.a .
van dam _
et al . _ ,
nature * 442 * , 667 ( 2006 ) .
y. asano , i. suemune , h. takayanagi and e. hanamura , phys .
rev .
lett * 103 * , 187001 ( 2009 ) .
see supplementary material for the details .
e. minot _ et al .
_ , nano lett .
* 7 * , 367 ( 2007 ) .
f. hassler _
et al . _ ,
nanotechnology * 21 * , 274004 ( 2010 ) . y .- m .
niquet and d.c .
mojica , phys .
rev .
b * 77 * , 115316 ( 2008 ) s.t .
yilmaz , p. fallahi and a. imamoglu , phys .
rev .
lett . * 105 * , 033601 ( 2010 ) .
m.o .
scully and w.e .
lamb , phys . rev . * 159 * , 208 ( 1967 ) .
j. heo , w. guo and p. bhattacharya , appl .
. lett .
* 98 * , 021110 ( 2011 ) .
p.b .
deotare _
et al . _ , appl .
. lett . * 94 * , 121106 ( 2009 ) .
m. notomi _
et al . _ ,
opt .
express * 16 * , 11095 ( 2008 ) .
s.e .
harris _
et al . _ ,
phys .
rev .
lett . * 18 * , 732 ( 1967 ) . |
asymmetry is an important feature of astrophysical dynamos and its properties may be important in understanding the specific processes . in the case of solar dynamo several asymmetric properties are well discussed , they may have important roles to maintain the dynamo process in spite of the cowling theorem .
the mirror asymmetry of the rotating vortices due to the coriolis force in the northern and southern hemispheres ( murakzy and ludmny , 2008 ) are important ingredients of the alpha effect .
deviations from the axial symmetry manifest themselves in a multipole structure by producing active longitudes ( berdyugina et al . , 2003 ) .
these are also important in the active stars ( olh et al . , 2009 ) .
the present work focuses on the solar north - south asymmetry with special emphasis on its variations .
the early works in this field are based on sunspot data ( newton and milsom , 1955 ) and reported no regular variations .
later works resulted in a diversity of temporal variations .
of course , the obtained different periods do nt exclude each other but further studies are necessary to clarify the most important types of variations .
we will return to these results in the discussions .
the observational bases of the work are the greenwich photoheliographic results ( gpr ) for cycles 12 - 20 , the kislovodsk sunspot data for cycle 21 and the debrecen photoheliographic data ( dpd , gyri et al . 2009 ) for cycles 22 - 23
. the combined dataset from different sources may be inhomogeneous from several aspects but here we focus on the north - south asymmetry and it is the same for images of any sources .
the institutional biases are the same for the two hemispheres and by applying any correction factors for both hemispheres the rate of asymmetry would not be distorted .
this assumption has also been checked by different homogenization procedures and we will demonstrate the source - independence of the results .
the first method to study the hemispheric phase lags is based on the cycle profiles .
the center of mass has been computed for each hemispheric cycle profile so that the height of this point represents the strength of the cycle and its temporal position shows the date to which the cycle as a whole can be bound .
this approach does nt consider the highly irregular shapes of the cycle profiles which could mislead the comparisons , in this method the bulk of the hemispheric cycle , i.e. the torus as a whole is considered regardless of the short - term fluctuations .
. shows the cycles 12 - 23 for the northern and southern hemisphere with the centers of mass .
figure 2 . shows the temporal differences between the centers of mass of both hemispheres .
the vertical axis shows the difference between the dates of northern and southern centers of mass so negative values mean that the northern cycle precedes .
it is conspicuous that the phase difference gradually changes from northern precedence to southern between cycles 12 - 19 and then , by an abrupt reversal , the southern hemisphere takes again the leading role and the trend starts again .
the figure shows two panels , the first one is based on the original data of the three data sources whereas the second one is calibrated in the following way .
a correction factor has been obtained for the kislovodsk data by using the greenwich - kislovodsk data ratios in their overlapping period and the procedure was applied for the debrecen data based on the kislovodsk - debrecen overlap .
the two figures are practically identical , the largest difference between two phase - lag data is two days , unperceivable in the plots .
the reason of the differences is that the multiplication of all data in a cycle with a correction factor may shift the date of the minimum between the neighboring cycles and of course this also shifts the date of the center of weight .
anyhow , one can conclude that the insitutional biases do not distort the rates of precedence between the two hemispheres .
the phase lags can also be studied by a different method .
the normalized north - south asymmetry index form sunspot group numbers ( sgn ) is : @xmath0 it is plotted for the period under study in fig.3 .
the ascending and descending phases of the cycles are indicated by grey and white stripes .
one could hardly recognize any regularity in the curve but if one plots the asymmetry index for the ascending and descending phases separately then the two curves of fig.4 .
are obtained .
the meaning of these curves is the same as that of fig.3 , as is demonstrated by the arrows connecting both points of each cycle : they have the same directions and lengths as those of fig.3 .
the explanation is that if e.g. the northern cycle precedes then the asymmetry index is higher in the ascending phase and lower in the descending phase ( as in the cycles 12 - 15 and 20 - 23 ) and the opposite is the case if the southern cycle precedes ( cycles 16 - 19 ) . figs 2 and 4 demonstrate the same phenomenon with two different methods .
the alternating hemispheric precedence , switching after 4 cycles , can be considered as a torsional oscillation of the solar global magnetic field .
the term torsional oscillation is primarily used for the fine structure of the differential rotation ( howard and labonte , 1980 ) , i.e. for a velocity field but it is also used for magnetic flux ropes , e.g. for those of sunspots ( gopasyuk , 2005 ) so the proper use of the term is not misleading . in the present case
the hemispheric initiating role seems to be alternating by 4 cycles .
this will deserve a closer scrutiny , here we only mention a possible further direction .
the advanced state of a cycle may also be related to the tilt angle of the active region magnetic axes ( connecting the leading and following parts ) .
fig.5 depicts the mean tilt angles for the northern and southern hemispheres in cycles 15 - 23 .
the data of the last two cycles are taken from the dpd ( gyri et al , 2009 ) the rest is taken from the mount wilson and kodaikanal tilt angle datasets ( howard et al , 1984 , sivaraman et al .
until cycle 19 ( period of southern precedence ) the northern angle is definitely higher whereas from cycle 20 ( period of northern precedence ) the values of the two hemispheres progress close to each other .
the only apparent exception is cycle 15 which has a small northern precedence and in fig.5 it seems to belong to the group of cycles 16 - 19 , when southern cycles precede .
anyhow , it may be conjectured that the hemispheric phase differences can be related to further signatures of the cycle progress such as the active region tilt angle or the hemispheric profiles of the differential rotation .
these questions need further study .
as was mentioned in the introduction several periods were published in earlier reports .
ballester et al .
( 2005 ) as well as zolotova and ponyavin ( 2007 ) published a 43-year peak in the asymmetry periodogram .
their methods are fourier analysis and cross - recurrence technique and they apparently may have found a harmonic of the period reported here .
the rest of the papers report fairly different periods .
vizoso and ballester ( 1990 ) reported 3.27 years , not detected later by ballester et al ( 2005 ) , they do not resolve the controversy .
chang ( 2007 ) published a period between 9 - 12 years , similarly to the 11.8 years period obtained by vizoso and ballester ( 1990 ) which is obviously the signal of the cycle .
as for the longer periods , braja et al .
( 2009 ) published a period of 70 years .
they admit , this is shorter than the gleissberg cycle and consequently from that published here .
the reason of the discrepancy may be that they simply focused on the time behaviour of asymmetry index without considering the individual cycles and their phases .
even longer periods were reported by forgcs - dajka et al .
( 2004 ) : longer than 100 years and by li et al .
( 2002 , 2009 ) : as long as 12 schwabe cycles . these periods are already too close to the length of the available datasets so it is difficult to interpret them .
two papers report most similar results to ours .
waldmeier ( 1971 ) also published an eight cycle long period of hemispheric phase - lags .
it is interesting that he followed a quite different method but obtained an almost identical behaviour to our results .
he computed the cyclic average of the yearly mean north - south latitudinal differences of sunspot groups , which is a quite different indicator of the hemispheric precedence than ours , and obtained a very similar time profile .
moreover , his period covers the cycles 10 through 20 and the cycles 10 - 11 have southern precedence and this corroborates the existence of the 8-cycle period presented here .
the other similar work was published by zolotova and ponyavin ( 2009 ) .
their method is also different from ours and that of waldmeier , e.g. , they consider the possibility of phase - lag sign reversal also within the cycle , namely at the maximum of cycle 16 .
this is the reason why their period is not bound to the length of eight entire cycles .
the presented feature , the alternating hemispheric precedence of the cycle progress over eight schwabe cycles , raises the question of the long - term variations of solar dynamo .
the self - consistent dynamo theories mainly deal with the reconstruction of the observed features of the individual cycles or two consecutive cycles and the central problem is the toroidal - poloidal transit in any types of models . the literature of solar dynamo is enormous , a recent review on the different types of dynamo models is given by jonas et al .
( 2009 ) with ample literature therein .
the authors show that all types of dynamo , even the most popular mean field models have serious difficulties in describing the way of building up the poloidal field .
the recently very popular forecast attempts encounter an even more difficult problem : some information should be retained by an unspecified solar memory for at least two cycles to be able to esteem the outlooks for the next activity , see e.g. solanki et al .
( 2002 ) and hathaway ( 2009 ) .
the above presented feature , however , would need a much longer memory which could ensure that after eight cycles the precedence should switch again from northern to southern type .
this seems to be fairly difficult to be incorporated into the recent dynamo models .
for this reason it may be timely to reconsider the possible role of certain external impacts on the dynamo regime as a modulating effect .
as an example , this may also be the case in seeking a plausible explanation for the long - term modulation of the cycle strength , in particular , the maunder minimum . following the work of jose ( 1965 ) and fairbridge and shirley ( 1987 ) , it is conspicuous that the inertial motion of the solar center of mass and the solar activity over several centuries have remarkable similarities .
of course , this does nt mean that the solar inertial motion would be the cause of the solar cycles but its modulating effect can not be excluded .
similarly , one can conjecture that the present variation could also be resulted by the heliographic latitudinal motion of the solar system barycenter , but this will also be addressed in a further study .
the research leading to these results has received funding from the european community s seventh framework programme ( fp7/2007 - 2013 ) under grant agreement no .
* ballester , j. l. , oliver , r. , carbonell , m. : 2005 , _ _ , * 431 * , l5 . * berdyugina , s. v. , usoskin , i. g. , 2003 , _ _ , * 405 * , 1121 .
* braja , r. , whl , h. , hanslmeier , a. , verbanac , g. , rudjak , d. , cliver , e. , svalgaard , l. , roth , m. : 2009 , _ _ , * 496 * , 855 . * chang , h - y . : 2007 , _ new ast .
_ , * 13 * , 195 . * fairbridge , r. w. , shirley , j. h. : 1987 , _ _ , * 110 * , 191 .
* forgcs - dajka , e. , major , b. , borkovits , t. : 2004 , _ _ , * 424 * , 311 . *
gyri , l. , baranyi , t. , ludmny , a. et al . , 2009 , debrecen
photoheliographic data for 1986 - 2003 , see : http://fenyi.solarobs.unideb.hu/dpd/index.html * gopasyuk , s. i. , gopasyuk , o. s. : 2005 , _ _ , * 231 * , 11 . *
hathaway , d. h. : 2009 , space sci .
rev . , doi 10.1007/s11214 - 008 - 9430 - 4 * howard , r. , labonte , b. j. : 1980 , _ _ , * 239 * , l33 . * howard , r. , gilman , p. , gilman , p. : 1984 , _ _ , * 283 * , 373 . *
jones , c. a. , thompson , m. j. , tobias , s. m. : 2009 , space sci .
, doi 10.1007/s11214 - 009 - 9579 - 5 * kislovodsk sunspot group reports 1954 - 2007 , + see : http://158.250.29.123:8000/web/soln_dann/ * jose , p. d. : 1965 , _ astron.j .
_ , * 70 * , 193 . * li , k. j. , wang , j. x. , xiong , s. y. , liang , h. f. , yun , h. s. , gu , x. m. : 2002 , _ _ , * 383 * , 648 .
* li , k. j. , gao , p. x. , zhan , l. s. : 2009 , _ _ , * 254 * , 145 . * murakzy , j. , ludmny , a. : 2008 , _ _ , * 486 * , 1003 . * newton , h. w. , milsom , a. s. : 1955 , , 115 , 398 . *
olh , k. , kollth , z. and 9 .
co - authors : 2009 , _ _ , * 501 * , 703 .
* royal observatory , greenwich , greenwich photoheliographic results , 1874 - 1976 , in 103 volumes , + see : http://www.ngdc.noaa.gov/stp/solar/ftpsunspotregions.html * sivaraman , k.r . ,
gupta , s.s .
, howard , r.f . : 1999 , _ _ , * 189 * , 69 .
see at : http://www.ngdc.noaa.gov/stp/solar/ftpsunspotregions.html * solanki , s. k. , krivova , n. a. , schssler m. , fligge , m. : 2002 , _ _ , * 396 * 1029 . * vizoso , g. , ballester , j. l. : 1990 , _ _ , * 229 * , 540 . *
waldmeier , m. : 1971 , , * 20 * , 332 . *
zolotova , n. v. , ponyavin , d. i. : 2007 , _
_ , * 243 * , 193 . * zolotova , n. v. , ponyavin , d. i. , marwan , n. , kurths , j. : 2009 , , * 503 * , 197 . | a long - term variation is detected in the north - south differences of the solar activity , the phase difference of hemispheric cycles alternates by four schwabe cycles .
this variability is demonstrated by two different methods .
it can be the related to the gleissberg cycle .
the variation of the sunspot group tilts is similar to the hemispheric phase lags on a shorter timescale .
some possible relations are mentioned . |
interactions play a major role in the evolution of galaxies , from the hierarchical growth of small systems at high redshifts to a range of mergers in the current universe ( e.g. , schweizer 2000 ) . even though models have been describing the nature of gravitational interactions between galaxies for more than 50 years
( e.g. , holmberg 1941 , toomre & toomre 1972 ) , much remains to be learned .
for example , debate continues over the nature of the interactions even in such well studied systems as m51 , where models have been proposed involving either single or multiple close passages of the companion ( toomre 1978 , salo & laurakainen 2000 , wahde & donner 2001 , dobbs et al .
2009 ) .
observational studies of interacting galaxies therefore continue to add value in providing empirical descriptions of how perturbed galaxies behave , in the process yielding better insights into their structures .
we chose the arp 104 system for an optical imaging study based on curiosity concerning its apparently extremely smooth optically visible tidal bridge ( see schombert et al .
1990 ; hereafter sws90 , smith et al .
2007 ) , which led to its inclusion in arp s ( 1966 ) _ atlas of peculiar galaxies_. this primarily stellar bridge connects ngc 5216 , the e galaxy southern member of the pair , to the northern disturbed disk galaxy , ngc 5218 , over a projected span of 38 kpc based on an assumed distance of 42 mpc ( 0.20 kpc arcsec@xmath0 ) to the system . beyond its optically visible features ,
arp 104 also has several intriguing physical characteristics , many of which are discussed by roche ( 2006 ) , cullen et al .
( 2007 ; hereafter c07 ) , and olsson et al .
most of the gas in the system is associated with ngc 5218 and its tidal debris . from their multiwavelength study ,
c07 infer a molecular gas mass of @xmath1 m@xmath2 in ngc 5218 , or more than double the hi mass of @xmath3 m@xmath2 for the entire arp 104 system ( see also olsson et al .
ngc 5218 thus is well fueled to support its current episode of intense star formation .
smith et al .
( 2007 ) further show that the thermal infrared emission is strong and centrally concentrated in ngc 5218 , while thermal infrared is much weaker but also centrally located in ngc 5216 .
this study , based on optical photometric imaging from the wiyn observatory 0.9-m and 3.5-m telescopes , focuses on the nature of the interactions responsible for arp 104 .
we summarize our observations in 2 and present the results in 3
. the astrophysical implications of these data are discussed in the context of the interaction and distribution of star formation in ngc 5218 in 4 , and a summary is provided in 5 .
imaging observations were made with the wiyn 0.9 m telescope located at the kitt peak national observatory in arizona on the nights of april 3 - 9 , 2008 .
we used the s2 kb ccd with an image area of 2048 x 2048 pixels of size 0.60 arcsec creating a field of view of 20.5 arcmin .
we observed with two standard harris filters , b and r. flat fields and bias exposures were taken at the beginning of each night . for each of our objects
, we took three exposures with each filter , 3 @xmath4 900 seconds in b and 3 @xmath4 600 seconds in r. each exposure was dithered from the original position so that ccd flat field deviations could be corrected for in the combining process .
the seeing disk full width at half maximum was 1.6 arcsec as estimated from stellar profiles , corresponding to a linear scale of 0.3 kpc .
the section of the combined r - band image containing arp 104 is shown in figure 1 .
the lower panel of figure 1 presents a deeper stretch of the image with a contour located at the 2.5 @xmath5 detection limit per seeing resolution element of @xmath625.5 mag arcsec@xmath7 . the signal - to - noise ratio in measurements of features with larger extent than the seeing disk scale with solid angle , @xmath8 , for the regime where flat field imperfections and scattered light do not introduce additional noise .
measurement precision can improve for objects subtending larger solid angles until flat field errors or other effects dominate , and thus no single detection limit exists for surface brightness studies of resolved objects ( e.g. , matthews & gallagher 1997 ) .
however , the contour in figure 1 provides a conservative limit to the reliability of our 0.9-m data .
arp 104 also was imaged with the wiyn 3.5-m telescope equipped with the minimosaic camera on march 19 , 2009 in the r - band , and a narrow band h@xmath9 filter .
the seeing was mediocre , @xmath101.2 arcsec .
exposure times for these two filters in the resulting combined images were 1800 s , and 5400 s , respectively .
these data were not photometrically calibrated , as they were designed to show the inner structures of ngc 5216 and ngc 5218 and detect hii region emission .
we also used these images to check for candidate young stellar complexes in the tidal bridge . due to the gap in the minimosaic camera ,
our coverage of the tidal bridge is incomplete , but no obvious candidate young stellar concentrations with high luminosity were found .
data were reduced using normal tasks in iraf to remove the bias , mask bad pixels , and flat field using dome flats .
cosmic rays were removed with the lacosmic routine ( van dokkum 2001 ) , and frames were aligned and stacked without additional geometric correction .
the photometric calibration from standard stars observed that night with the 0.9-m telescope proved unreliable .
we therefore transformed the 0.9-m images to the sdss _
g _ and _ r _ ab magnitude system using stars in the field of view .
the resulting transformations are linear in _ r _ with a small color term .
the systematic transformation errors are estimated to be @xmath110.07 mag in _
r _ and @xmath110.05 mag in _
g _ @xmath12 _
r _ from the scatter around the fit .
the iraf tasks apphot and ellipse yielded aperture photometry and isophote fitting on the calibrated 0.9-m images and are illustrated in figures 3 and 5 .
determining colors of the tidal bridge is difficult due to its low surface brightness .
however , measurements of the bridge surface brightness show that it is remarkably uniform , varying by @xmath1110% along the ridgeline between the two galaxies . colors of the bridge therefore could be determined by using rectangular apertures along the ridgeline .
a g@xmath12r color map is shown in figure 2 , and photometric measurements are combined with a summary of other basic information in table 1 . a check of our g - band photometry transformed to b following jordi et al .
( 2006 ) against the b - magnitude listed in the nasa extragalactic database ( ned ) yields agreement at the 0.1 mag level .
our results also are generally consistent with those of sws90 in that we find the bridge to be moderately blue , although we derive somewhat brighter magnitudes for the two main galaxies .
we are unable to confirm the reality of the outer blue regions which they found around ngc 5216 and the bridge .
the objectives of the wiyn 3.5-m imaging were to observe hii regions and map smaller angular scale structures .
the central parts of both galaxies are illustrated in figure 3 along with narrow band images of ngc 5218 made by subtracting an appropriately scaled r - band image as an off - band .
intense h@xmath13[nii ] line emission is present in both nuclei ( zhao et al .
2006 ) , and partially accounts for their red g - r colors .
dust lanes also are obvious in the central regions of ngc 5218 , consistent with the presence of a larger central concentration of molecular gas and associated strong radio continuum source ( c07 , olsson et al .
the structural classification of ngc 5216 as an elliptical galaxy is supported by its mean radial brightness profile which is well fit by an @xmath14 law from a radii of 3 to @xmath1030 arcsec , i.e. to about 6 kpc ( figure 4 where an r@xmath15 profile will be a line ) . beyond this the presence of shells and the tidal tail produce deviations in the profile , even though the overall trend remains close to an r@xmath15 intensity scaling out to the last reliable data point located slightly beyond r@xmath1618 kpc .
the surface photometry was carried out using the ellipse task in the iraf package stsdas , with the center position held fixed and the position angle free to be adjusted .
this approach also was used to derive the radial brightness distribution of ngc 5218 presented below . beyond the possibly dusty nucleus with its emission lines ,
the _ g_@xmath12_r _ color of ngc 5216 also is consistent with a typical elliptical galaxy and it then is not surprising to find that this system lies on the red sequence .
its liner nuclear spectrum ( zhao et al . 2006 ) also is a relatively common feature in an elliptical galaxy .
however , for radii of r@xmath176 kpc , the brightness profile of ngc 5216 flattens , and our results are consistent with those of roche ( 2006 ) . in this region shells exist which can be seen in the wiyn 3.5-m images ( figure 5 ) along with the presence of the southern tidal tail that extends across the galaxy ( figure 3 ) .
the tidal tail is the only region with confirmed recent star formation as indicated by its blue colors .
the outer shells in ngc 5216 unfortunately are too faint , @xmath1825 mag arcsec@xmath7 , for us to obtain reliable color measurements from our data , although such measurements would be useful in constraining when the shells formed in ngc 5216 ( e.g. , mcgaugh & bothun 1990 ) .
ngc 5218 , the northern member of the arp 104 pair , is well known to have a disturbed structure .
its fundamental characteristics as a disk galaxy , however , are clear from the optical structure and from results of the radial brightness profile determined from isophotal surface photometry shown in figure 6 as well as from surface photometry in roche ( 2006 ) .
a simple interpretation of figure 6 is to consider it as indicating the presence of two approximately exponential disks .
the inner disk then has a scale length of @xmath109 arcsec ( 2 kpc ) while the outer disk scale length increases at radii @xmath1775 arcsec where the luminosity is increasingly dominated by the tidal debris , which have a shallower mean radial intensity profile than that of the outer disk . a high surface brightness central linear feature with a length of @xmath1915 arcsec ( 3 kpc ) , possibly a bar as suggested by c07 and the kinematics found by olsson et al .
( 2007 ) , or an edge - on nuclear disk , is seen at pa@xmath1095@xmath20 . beyond the central linear feature
is an ` inner disk ' with major axis at pa@xmath1060@xmath20 and with a radius of 7 kpc .
optical indications for dust are seen in this region and become prominent in the center of the galaxy , consistent with the intense central far infrared emission observed by smith et al .
( 2007 ) and the presence of substantial molecular gas concentrations ( cullen et al .
2007 , olsson et al . 2007 ) .
dust therefore can be expected to influence optical colors in the inner parts of ngc 5218 , but there are no indications for substantial amounts of dust in the outer disk or tidal material .
olsson et al .
further conclude that the nucleus appears to be a starburst - driven liner based on its radio continuum and molecular line properties . the inner disk is encompassed by an ` outer disk ' with major axis pa@xmath1090@xmath20 that extends to r@xmath1015 kpc .
figure 2 shows that the two main disks also have different colors , with the outer structures being bluest , with colors indicating star formation occurred throughout much of the galaxy at significant levels in the past 10@xmath21 yr .
roche ( 2006 ) also found spectroscopic evidence for moderately young stars in ngc 5218 from his integrated spectra which displayed strong h - balmer absorption lines from an intermediate age stellar component . while the integrated color and absolute magnitude of ngc 5218 place it in the green valley " , this is not necessarily a proper representation of the global stellar population in this galaxy . instead the integrated color reflects a wide range of local conditions , extending from bluest colors in the outer shell regions where relative star formation rates currently are low to red colors in the dusty central high star formation rate zone . the locations of very recent star formation ( last @xmath1010 myr ) traced by the h@xmath13[nii ] emission support this picture
the majority ( @xmath1960% ) of the h@xmath13[nii ] flux occurs in and around the central linear structure , which also is the location of the luminous thermal infrared emission ( smith et al .
the remainder is distributed primarily along the rim of the inner disk , with a few hii regions detected in the outer disk .
the presence of hii regions indicates that the reddening of g@xmath12r color toward the center of ngc 5218 is associated with the distribution of dust ( see figure 3 and the far infrared maps of smith et al .
( 2007 ) ) rather than an absence of young stellar populations .
in particular the very red nuclear color that we observe apparently is fostered by a combination of dust and intense h@xmath13[nii ] emission , and follows the tendency for dusty central star formation seen in samples of green valley galaxies ( cowie & barger 2008 , brammer et al .
2009 ) .
the impression that the two main disks are tilted with respect to one another is supported by the kinematic data presented in c07 .
they note that the inner velocity field as measured from the co 1 - 0 transition appears to rotate counter to that defined by the more radially extended hi ( figure 3 ) .
this can be understood as result of a strongly warped disk component that changes in angle with radius ( e.g. sparke et al .
we therefore conclude that the disk system in ngc 5218 is unlikely to be coplanar , but instead shows the large range in angular momentum vector direction that is typical of galaxies which have experienced recent mergers ( barnes 2002 ) .
the combination of structures and kinematics in ngc 5218 is reminiscent of the out - of - plane features seen in polar ring galaxies , such as ngc 3718 ( sparke et al .
2009 ) , and suggests that ngc 5218 is the product of a relatively recent merger . as a result of this event
, the optical structure of ngc 5218 is still evolving .
in particular the central concentration of star formation along with its high dust content compromise our ability to determine the ultimate bulge - to - disk ratio of the merger remnant from its current optical structure ( see lotz et al .
thus we can not tell from optical surface photometry alone if ngc 5218 is a product of a major merger of near equal mass galaxies , or due to some other class of minor merger event .
measurement of the stellar velocity field will be helpful in revealing the inner structure of ngc5218 and thus clarifying this issue .
for example , kinematic data , especially if obtained in the near - infrared , can reveal whether the central linear feature is likely to be a bar observed at moderate inclination or a possibly a combination of bar and highly inclined central disk ( c.f .
olsson et al . 2007 ) , and whether a significant spheroidal bulge is present . following the discussion of c07 and using the results from surace et al . ( 2004 ) , ngc 5218 has sfr @xmath105 - 7 m@xmath2 yr@xmath0 .
the gas exhaustion time scale with the current sfr is @xmath191.3 - 1.9 gyr , or longer than the likely time for the two systems to merge ( see also c07 ) .
the stellar bridge between galaxies is well resolved with a the width of @xmath1910 arcsec ( 2.2 kpc ) , and _
r_-band luminosity of only @xmath192% of that of ngc 5218 .
it also is remarkably constant in surface brightness . combining the 0.9-m telescope b- and
r - band images in an unweighted average to gain signal - to - noise , the bridge surface brightness determined in [email protected] arcsec@xmath22 regions remains constant to @xmath2310% .
the only indication of small scale structure is a marginally detected decrease in surface brightness by @xmath1920% in the vicinity of the hi emission peak mapped by c07 .
the structure perpendicular to the long axis of the bridge is similarly regular , with a slow rise in intensity from the east and a sharp drop off to the west after the intensity peak .
the form of the spatial gradient in surface brightness perpendicular to the long axis of the bridge is consistent with its origin in a prograde interaction between ngc 5216 and ngc 5218 ( oh et al .
2008 ) , in agreement with the basic features of the c07 ngc 5216-ngc 5218 interaction model .
we estimated the minimum stellar mass density along the ridge of the tidal bridge by adopting a single age galev stellar population model ( kotulla et al .
2009 ) with a kroupa - imf and slightly sub - solar metallicity appropriate to an outer galactic disk .
minimum _ stellar mass was found by assuming the bridge consists of a 1.1 gyr age simple stellar population which would have _
@xmath10 0.5 .
note that a more realistic model is likely to contain older stars and thus have a higher mass - to - light ratio , but even with only a _
g_@xmath12_r _ color to constrain the stellar population parameters , this simple model provides an approximate lower limit to the stellar mass density .
this match gives @xmath2420 m@xmath2 pc@xmath7 , and a mean stellar density of @xmath250.01 ( @xmath26 kpc ) m@xmath2 pc@xmath27 where @xmath28 is the depth of the bridge along the line of sight .
these results indicate that the bridge is not rich in hi . the mean value of n(hi)@xmath1710@xmath29 cm@xmath7 required to equal the stellar density readily exceeds the c07 measured peak value of n(hi)@xmath1010@xmath31 cm@xmath7 .
we therefore conclude that @xmath32 0.1 along the bridge .
optical bridges between galaxies originate from stars which tides strip from outer galactic disks of interacting galaxies ( e.g. , toomre & toomre 1972 , oh et al .
the optical color of the arp 104 bridge is constant at g@[email protected]@xmath330.1 , and , in agreement with sws90 , the bridge color is consistent with its stellar populations originating in the outer disk of ngc 5218 ( see table 1 and roche ( 2006 ) ) .
we find no evidence for substantial levels of ongoing star formation along the bridge between galaxies , but can not exclude the presence of small star forming regions , such as those seen in the m81 tidal bridge ( de mello et al .
2008 ) .
the northern and southern optical tidal tails , are much shorter than the bridge , with less regular structures . based on the _ g_@xmath12_r _ color , significant young stellar populations with ages of @xmath3410@xmath21 yr are found in the northern tidal tail , and , perhaps surprisingly , also in its southern counterpart that extends from ngc 5216 .
the northern optical tidal feature is located at the base of the long hi tidal tail that extends to the northeast of ngc 5218 , most of which is not detected in the optical .
arp 104 therefore shows a highly asymmetric set of tidal features .
the majority of the hi gas lies in ngc 5218 and the northern tidal tail , while most of the tidally stripped stars are in the inter - galaxy bridge .
arp 104 is the dominant member of a small group of galaxies , lgg 854 ( garcia 1993 ) .
the other cataloged member of lgg 854 is ngc 5216a , an edge - on late - type disk galaxy at a projected distance of @xmath35580 kpc .
a search of ned also yields ugc 8436 at @xmath35 680 kpc and ugc 8491 at @xmath35230 kpc , both moderate luminosity late - type galaxies , and possible group members . a search of the 0.9-m images yielded only two lsb objects , but given the distance to arp 104 and image quality of the 0.9-m data , these can not be reliably classified .
however , we find no luminous dwarfs with m@xmath3615 within the 240 kpc field covered by the 0.9-m telescope images .
the combination of our optical measurements of the stellar structure and colors of arp 104 and multi - wavelength studies of roche ( 2006 ) and c07 indicate that arp 104 is the product of an interesting sequence of astrophysical processes : * as described by c07 , a prograde close passage between the disky ngc 5218 galaxy and the elliptical ngc 5216 occurred about 300 myr in the past .
the smooth stellar tidal bridge is a product of that interaction and comparisons with models show the two galaxies are likely to be separating .
the shearing along the tidal bridge as the two galaxies separate , along with its low stellar density , may explain the near constant surface brightness , since gravitational instabilities will have difficulty in growing under these conditions .
as discussed by oh et al .
( 2008 ) , the surface brightness of tidal bridges increase with interaction strength , so the low brightness of the arp 104 bridge then suggests that ngc 5218 did not make an extremely close passage by ngc 5216 .
* the tidal tails in the system include features with hi gas and younger stellar populations extending to the southwest from ngc 5216 with similar conditions present at the base of primary hi tail running to the north and east of ngc 5218 . the northern tidal tail , however , is optically faint and therefore stands in contrast to the bridge joining the two galaxies , where stars dominate .
the origin of this asymmetry is not immediately obvious in the context of a purely binary galaxy interaction model .
our color estimates and those of sws90 for the tidal bridge are consistent with its stars coming from ngc 5218 , and are insufficiently blue for the bridge to consist primarily of stars formed during the interaction . *
ngc 5218 contains stellar shells with the bluest optical colors in the galaxy , showing that they hosted significant star formation within the past @xmath191 gyr .
these features are not reproduced by the interacting galaxy model of c07 .
stellar shells usually are associated with mergers involving disk galaxies ( e.g. , schweizer & seitzer 1988 ) , which raises the question as to whether ngc 5218 experienced a merger before interacting with ngc 5216 . two possibilities are that this material came from an interaction with ngc 5216 , or involved a minor merger of ngc 5218 with a third galaxy . given the red sequence colors , elliptical galaxy structure without an obvious disk , and apparent lack of a cool interstellar medium in ngc 5216 , we favor the recent minor merger option for the origin of shells in ngc 5218 .
a previous merger could have broken the axial symmetry of ngc 5218 , which could lead to the currently observed structural differences in the tidal tails .
modeling is needed to explore this possibility . *
the distribution of ` disk ' hii regions in two ellipsoidal rings inclined to each other also suggests that ngc 5218 experienced a significant external perturbation .
while a previous encounter with ngc 5216 might explain the situation , the counter rotation observed in the @xmath37co line by c07 is more readily achieved by material associated with a previous merger ( bureau & chung 2006 , vergani et al .
2007 , crocker et al .
2009 and references therein ) . * the integrated colors of ngc 5218
place it in the green valley , between the blue plume and red sequence galaxies . in agreement with other recent studies ( e.g. , cowie & barger 2008 , brammer et al . 2009 ) , this color reflects a combination of blue colors from young but aging stellar populations and a dusty central region .
if a future merger were not a factor , then ngc 5218 contains sufficient gas to fuel star formation at a rate that would allow it to remain in something close to its current state for up to @xmath191 gyr . *
ngc 5216 also displays faint outer shell structures .
their presence complicates the interpretation of this system .
are these remnants of yet another previous interaction ? or could these features result from material captured from ngc5218 ?
if so we would expect them to have blue optical colors , typical of the outer regions of ngc5218 .
unfortunately we were not able to make the necessary measurements from our data . *
both galaxies host energetic processes in their nuclei .
ngc 5216 has a liner optical spectrum ( zhao et al .
2006 ) and the h@xmath13[nii ] emission probably is a major source of the very red nuclear colors that we observe .
optically visible star formation , the radio continuum ( c07 , olsson et al .
2007 ) and thermal infrared emission ( smith et al .
2007 ) is strongly concentrated within the central linear feature of ngc 5218 , which produces most of its h@xmath13[nii ] emission line luminosity .
since this region also contains obvious dust lanes and and a dense concentration of molecular gas ( olsson et al . 2007 ) , this is a lower bound on the h@xmath9 luminosity fraction , which will increase when corrected for internal extinction .
the multi - wavelength data are consistent with the presence of a nuclear starburst and possible liner in ngc 5218 . *
multiple mergers have been suggested to be factors in early - type galaxies involved in interactions in galaxy groups ( e.g. , prez grana et al .
thus arp 104 potentially illustrates a significant evolutionary pathway for the production of early - type galaxies in small groups . in this case
we find products of possible merging sequences where ngc 5218 combined with another system a few internal dynamical times ago and probably will merge with ngc 5216 a few internal dynamical times in the future .
this process contrasts with the possibility of rapid multiple mergers , e.g. , in which more than two members of a compact galaxy group might combine to form one galaxy on approximately a compact region crossing time or few internal galactic dynamical times ( e.g. , rubin et al . 1990 ,
amram et al .
2007 ) .
based on standard galaxy interaction models , such as that in c07 , the eventual fate of arp 104 will be to merge into a single galaxy , as also suggested by roche ( 2006 ) .
whether a sufficient reservoir of gas with high specific angular momentum will remain to allow a star - forming disk to exist in the resulting merged system is unclear .
while the current supply of gas ( @xmath1010@xmath38 m@xmath2 ) is adequate to make a moderately luminous stellar disk , its eventual state depends on the properties of gas in the tidal tails in combination with the effectiveness of processes in funneling gas into the central region where high densities promote dissipation and star formation and can lead to building a bulge ( cf .
hopkins et al .
2008 ) .
although standard spiral - spiral merger models predict that only a small disk is likely to remain once a merged system settles into its final form over @xmath190.3 - 1 gyr ( e.g. , lotz et al .
2008 , hopkins et al .
2009b ) , this conclusion may not apply to major spiral - elliptical mergers , such as that in the future for arp 104 ( di matteo et al .
furthermore , a sequence of mergers closely spaced in time , as we are suggesting for arp 104 , could affect the angular momentum of gas in the tidal tails , and therefore the structure of the final merged system .
thus the galaxy resulting from the ngc 5216-ngc 5218 major merger might have either an approximately sa - type morphology if some of the ngc 5218 disk is retained and renewed through post - merger star formation , or an s0/e structure if dissipation dominates and gas primarily collects in the central region of the merged galaxy ( bekki 2001 , hopkins et al .
2009a ) .
as the group containing arp 104 at most contains a few widely scattered and much less luminous members , this system could be headed towards becoming what appears to be an isolated elliptical galaxy ( aars et al .
however , it will have achieved this state through successive mergers , and thus the isolation would be a product of environmental `` nurture '' in a small galaxy group , rather than `` nature '' in the form of initial conditions leading to the formation of a relatively massive early - type galaxy at high redshift ( e.g. bournaud et al . 2007 ) .
our study of the optical structure of arp 104 reveals features which illuminate the evolutionary state of this system .
the presence of shells in both members of the pair and apparent counter - rotation in ngc 5218 suggest that they both have been influenced by interactions .
these features possibly could result from earlier encounters between the two galaxies , or , given the angular momentum misalignment in ngc 5218 , in our view more likely from its involvement in an earlier merger .
this also would have distorted the system and therefore might be the origin of the asymmetry in the tidal bridges , with hi dominating to the north and stars between ngc 5216 and ngc 5218 .
if this idea is correct , then the structure of the ultimate merger of the arp 104 system also could be affected , e.g. , if infalling tidal material were to have gained sufficient angular momentum to avoid being incorporated into a bulge and instead aiding the survival of a disk component .
the disturbed structure of ngc 5218 suggests interactions moved gas inwards , leading to star formation becoming centrally concentrated in a dusty , molecular gas - rich nuclear starburst zone .
the integrated optical colors place this galaxy in the green valley , while ngc 5216 is on the red sequence .
however , consistent with other recent statistical studies , the green valley color of ngc 5218 stems from a wide range of recent star formation histories and does not properly represent the current state of the galaxy , which supports intense star formation in its dusty central regions .
while deeper imaging would be useful , the lack of bright dwarfs or tidal debris differing from those readily associated with the binary pair arp 104 appears inconsistent with expectations if this system were the late stage product of dynamical timescale merging in compact galaxy group .
we therefore favor a hierarchical merger model where interactions first led to ngc 5216 and ngc 5218 dominating their galaxy group , with the future product being a merger into a single luminous sa or earlier type galaxy that we now catch in the process of formation . this first phase of this project comprised a. parker s senior thesis in astronomy at the university of wisconsin - madison .
we thank the wiyn observatory teams , especially hillary mathis of the wiyn 0.9-m obervatory , for making it possible to obtain the observations reported here , kate dellenbusch for her help with the wiyn 0.9-m observations , and an anonymous referee for useful comments and for pointing out the unpublished very interesting study of arp 104 by n. roche .
jsg benefited from discussions with susanne aalto and evert olsson during a visit to the onsala space observatory .
he also thanks the university of canterbury , christchurch , new zealand for the opportunity to hold an erskine visiting fellowship during the revisions of this manuscript .
this research was supported in part by the national science foundation through grant ast-0708967 to the university of wisconsin and has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
aars , c. e. , marcum , p. m. , & fanelli , m. n. 2001 , , 122 , 2923 amram , p. , mendes de oliveira , c. , plana , h. , balkowski , c. , & hernandez , o. 2007 , , 471 , 753 arp , h. 1966 , , 14 , 1 barnes , j. e. 2002 , , 333 , 481 bekki , k. 2001 , , 547 , 189 bournaud , f. , jog , c. j. , & combes , f. 2007 , , 476 , 1179 brammer , g. b. , et al .
2009 , , 706 , l173 bureau , m. & chung , a. , 366 , 182 cowie , l. l. , & barger , a. j. 2008 , , 686 , 72 crocker , a. f. , jeong , h. , komugi , s. , combes , f. , bureau , m. , young , l. m. , & yi , s. 2009 , , 393 , 1255 cullen , h. , alexander , p. , green , d. a. , & sheth , k. 2007 , , 376 , 98 di matteo , p. , combes , f. , melchior , a .-
l . , & semelin , b. 2008 , , 477 , 437 dobbs , c. l. , theis , c. , pringle , j. e. , & bate , m. r. 2010 , , 403 , 625 garcia , a. m. 1993 , , 100 , 47 holmberg , e. 1941 , , 94 , 385 hopkins , p. f. , hernquist , l. , cox , t. j. , dutta , s. n. , & rothberg , b. 2008 , , 679 , 156 hopkins , p. , cox , t. j. , younger , j. d. , & hernquist , l. 2009a , , 691 , 1168 hopkins , p. f. , sommerville , r. s. , cox , t. j. , hernquist , l. , jogee , s. , keres , d. , ma , c .- p . , robertson , b. , & stewart , k . 2009 , , 397 , 802 rubin , v. c. , hunter , d. a. , & ford , w. k. , jr . 1990 , , 365 , 86 jordi , k. , grebel , e. k. , & ammon , k. 2006 , , 460 , 339 kotulla , r. , fritze , u. , weilbacher , p. , & anders , p. 2009
, , 396 , 462 lotz , j. m. , jonsson , p. , cox , t. j. , & primack , j. r. 2008 , , 391 , 1137 matthews , l. d. , & gallagher , j. s. , iii 1997 , , 114 , 1899 mcgaugh , s. s. , & bothun , g. d. 1990 , , 100 , 1073 de mello , d. f. , smith , l. j. , sabbi , e. , gallagher , j. s. , mountain , m. , & harbeck , d. r. 2008 , , 135 , 548 neff , s. g. , et al .
2005 , , 619 , l91 oh , s. h. , kim , w .- t . , lee , h. m. , & kim , j. 2008 , , 683 , 94 olsson , e. , aalto , s. , thomasson , m. , beswick , r. , & httemeister , s. 2007 , , 473 , 389 prez grana , j. a. , kemp , s. n. , de la fuente , e. , katsiyanna , a. c. , & meaburn , j. 2009 , new astronomy , 14 , 556 roche , n. 2006 , arxiv : astro - ph/0605015 salo , h. , & laurikainen , e. 2000 , , 319 , 377 schombert , j. m. , wallin , j. f. , & struck - marcell , c. 1990 , , 99 , 497 schweizer , f. 2000 , phil . trans .
london , 358 , 2063 .
schweizer , f. , & seitzer , p. 1988
, , 328 , 88 smith , b. j. , struck , c. , hancock , m. , appleton , p. n. , charmandaris , v. , & reach , w. t. 2007 , , 133 , 791 sparke , l. s. , van moorsel , g. , schwarz , u. j. , & vogelaar , m. 2009 , , 137 , 3976 surace , j. a. , sanders , d. b. , & mazzarella , j. m. 2004 , , 127 , 3235 toomre , a. , & toomre , j. 1972 , , 178 , 623 toomre , a. 1978 , in large scale structures in the universe , symp .
estonian ssr , 109 van dokkum , p. g. 2001 , , 113 , 1420 vergani , d. , pizzella , a. , cosini , e. m. , van driel , w. , buson , l. m. , dettmar , r .- j . , & bertola , f. 2007 , , 463 , 883 wahde , m. , & donner , k. j. 2001 , , 379 , 115 zhao , y .- h .
, gu , q .-
s . , peng , z .- x . ,
shi , l. , luo , x .-
l . , & peng , q .- h .
2006 , chinese journal of astronomy and astrophysics , 6 , 15 ccllc object & region & _ r _ or @xmath39 & _ g _ @xmath12 _ r _ & comment + ngc 5216 & integrated & [email protected] & [email protected] & [email protected] + & envelope & & [email protected] + ngc 5218 & integrated & [email protected] & [email protected] & [email protected] + & nuclear zone & & [email protected] & average of 2 sides + & outer arms & & [email protected] & + north tidal tail & outer , r@xmath4160 arcsec & @xmath4223.8 & [email protected] & + bridge & & @xmath624.2 & [email protected] & constant @xmath39 to @xmath1010% + south tidal tail & r=55 arcsec & @xmath4223.6 & [email protected] & + 110 kpc .
the lower image shows a deeper stretch with the contour located at the 2.5 @xmath5 detection level for an object with the 1.6 arcsec seeing disk full width at half maximum , corresponding to 0.3 kpc .
the bar in the lower panel is 10 kpc in length .
[ arp104_r1 ] ] _ r _ color map from observations with the wiyn 0.9-m telescope ; n to left and e is down .
the field of view is the same as in figure 1 .
the central part of the tidal bridge is at the low surface brightness limit for reliable _ g_@xmath12_r _ color measurements with these data .
[ arp104cols ] ]
redshifted with respect to central velocity , @xmath12 blueshifted ) from c07 ( upper left ) ; h@xmath13 [ nii ] narrow band ( upper right ) ; ngc 5218 high pass image showing the dust lanes and outer shells ( lower right ) .
ngc 5216 high pass spatially filtered wiyn 3.5-m r - band image with stellar intensity contours showing the southwestern tidal tail extending from the center of ngc 5216 ( lower left ) .
[ arp104_35 ] ] -law profile typical of an elliptical galaxy .
the actual data are very close to the r@xmath15 profile in the inner regions , but display significant deviations at larger radii associated with the outer shells .
error bars from the ellipse fitting task are plotted ( in color online for increased visibility ) when they exceed 0.05 mag .
r@xmath433 arcsec@xmath15 corresponds to [email protected] kpc .
[ n5216_isophot ] ] . at r@xmath1775
arcsec the presence of tidal debris make surface photometry of the disk increasingly unreliable .
the line shows an exponential disk model with a scale length of 2 kpc and central surface brightness offset by @xmath442.4 mag arcsec@xmath7 for visibility from the the mid - disk model fit of @xmath4518.6 mag arcsec@xmath7 .
[ n5216_deepr ] ] | arp 104 is a pair of luminous interacting galaxies consisting of ngc 5216 , an elliptical , and ngc 5218 , a disturbed disk galaxy and joined by a stellar bridge .
we obtained optical imaging to support photometric and color studies of the system .
ngc 5216 lies on the red sequence , while the unusual distribution of stellar population properties in combination with intense central star formation in a dusty region result in ngc 5218 being a nearby example of an intermediate color ( green valley ) system .
the stellar bridge has remarkably uniform optical surface brightness , with colors consistent with its stars coming from the outskirts of ngc 5218 , but is relatively gas - poor while the northern tidal tail is rich in hi . while both galaxies contain shells , the shell structures in ngc 5218 are pronounced , and some appear to be associated with counter - rotating gas .
this combination of features suggests that arp 104 could be the product of distinct multiple interactions in a small galaxy group , possibly resulting from a hierarchical merging process , and likely leading to the birth of a relatively massive and isolated early - type galaxy .
_ astrophysical journal , in press , august 2010 _ |
the high mass x - ray binary source hmxrb source oao 1657 - 415 ( oao 1653 - 40 ) was first detected by the copernicus satellite ( polidan et al .
1978 ) in the 4 - 9 kev range .
the heao-1 observations also showed 38.22 sec pulsations in the 1 - 40 kev and 40 - 80 kev bands ( white @xmath1 pravdo 1979 , byrne et al .
observations with ginga and granat ( kamata et al .
1990 , gilfanov et al .
1991 , mereghetti et al .
1991 , sunyaev et al .
1991 ) have found pulse period changes .
batse observations of this source with the compton gamma ray observatory ( cgro ) showed that oao 1657 - 415 is in an @xmath2 binary orbit with an x - ray eclipse by the a stellar companion ( chakrabarty et al .
the observed orbital parameters imply that the companion is a supergiant of spectral class b0-b6 .
the correlations between x - ray flux and pulse frequency derivatives ( @xmath3 ) fluctuations were investigated by using the previously published pulse frequencies and batse measurements ( baykal 1997 ) .
these correlations suggested that the formation of episodic accretion disks in the case of a stellar wind is the possible accretion mechanism . in this paper , we present the short term pulse frequency fluctuations and x - ray fluxes of oao 1657 - 415 in the light of recent rxte observations .
we have employed background subtraction by using the background models for the rxte / pca instrument and galactic ridge emission in the 2 - 50 kev range . our x - ray flux and pulse frequency measurements find an increase in the x - ray flux which is correlated with the decrease in the spin - down rate ( or marginal spin - up trend ) .
oao 1657 - 415 was observed between 1997 august 20 - 27 within the guest observer program of rxte with proposal observation i d 20113 .
rxte pointings of the source are separated from each other by approximately six hours with a total observation span of 75 ksec .
the results presented here are based on data collected with the proportional counter array ( pca , jahoda et al . , 1996 ) and the high energy x - ray timing experiment ( hexte , rothschild , et al . , 1998 ) .
the pca instrument consists of an array of 5 proportional counters operating in the 260 kev energy range , with a total effective area of approximately 7000 @xmath4 and a field of view @xmath5 fwhm .
the hexte instrument consists of two independent clusters of detectors , each cluster containing four nai(t1)/csi(na ) phoswich scintillation counters sharing a common @xmath5 fwhm field of view . the field of view of each cluster is switched on and off source to provide background measurements .
the net open area of the seven detectors is 1400 @xmath4 and each detector covers the energy range 15250 kev .
background light - curves and pulse height amplitudes are generated by using the background estimator models based upon the rate of very large events ( vle ) , spacecraft activation and cosmic x - ray emission with the standard pca analysis tools .
the background light - curves are subtracted from the source light - curve obtained from the good xenon event data ( fig .
since the source is close to the galactic center , galactic ridge data were extracted from archival rxte observations .
observations pointed at directions a few degrees away from the source are collected .
the background spectra for the galactic ridge data were generated using pca analysis tools .
after the instrument background are removed from the galactic ridge data , the residual spectra are used as a background for the x - ray spectra of oao 1657 - 415 .
the source spectrum was calculated using the same pca background estimator models .
( it should be noted that @xmath6 @xmath7 systematic errors are used in spectral fitting of pca data ) . for the hexte data the background subtraction is straightforward since the hexte detectors are rocking in 16 sec intervals .
standard ftools software for rxte is used for the data reduction and for the dead time correction .
we have used power - laws with high energy cut - off models together with a 6.65 kev gaussian emission line which was found in previous ginga observations ( kamata et al . ,
we have found no evidence for any deviation from a power - law model with high energy cut off which might be attributable to a cyclotron feature .
table 1 presents the spectral parameters of x - ray spectra .
figure 2 presents the joint x - ray spectra of rxte / pca and hexte detectors . the power - law index , and the cut - off and e - folding energies are significantly different from those measured by ginga , the powerlaw steeper , but extending to higher energy . the broader energy band of the pca and hexte combination better constrain the higher energy spectra , but the spectrum may vary intrinsically .
the background subtracted light - curves are corrected with respect to the barycenter of the solar system . using the binary orbital parameters of oao 1657 - 415 from batse observations ( bildsten et al . , 1997 ) , the light curves are also corrected for binary motion of oao 1657 - 415 ( see table 2 ) . a long power spectrum was used to estimate the average pulse frequency .
this pulse frequency is consistent with batse pulse frequency records at the same time ( obtained through heasarc http : cossc.gsfc.nasa.gov ) . in order to resolve the pulse arrival times and pulse frequencies at
shorter time scales 29 pulse arrival times were generated ( one pulse arrival time for each rxte orbit ) .
pulse arrival times are found by folding the light - curve data into one average pulse for each rxte orbit , folding all light - curves into one master pulse , and cross - correlating the master pulse with each of the 29 average pulses . in the pulse timing anaysis , we have used the method of harmonic representation of pulse profiles , as proposed by deeter @xmath8 boynton ( 1985 ) . in this method , pulse profiles are expressed in terms of harmonic series and cross - correlated with the master pulse profile .
the maximum value of the cross - correlation is analytically well defined and does not depend on the phase binning of the pulses .
the master pulse with 40 phase bins was represented by their harmonics ( deeter @xmath8 boynton 1985 ) and cross - correlated with harmonic representations of pulse profiles from segments of the data .
the pulse profiles for oao 1657 - 415 are variable .
this affects the pulse timing . in order to estimate the errors in the arrival times , the light - curve of each rxte orbit
is divided into approximately 10 - 15 equal subsets and new arrival times are estimated .
the average variance in the arrival times are computed and treated as errors of arrival times .
the pulse arrival times are represented in fig .
3 . the residual pulse arrival times may arise from the change of the pulse frequency during the observation ( or intrinsic pulse frequency derivative ) and from the errors of obital parameters ( deeter et al . , 1981 ) , @xmath9 where @xmath10 is the pulse phase offset deduced from the pulse timing analysis , @xmath11 is the mid - time of the observation , @xmath12 is the phase offset at t@xmath13 , @xmath14 is the deviation from the mean pulse frequency ( or additive correction to the pulse frequency ) , and @xmath15 is the pulse frequency derivative of the source , @xmath16 is the light travel time for projected semimajor axis , @xmath17 is the epoch when the mean orbital longitude is equal to 90 degrees , @xmath18 is the orbital period , @xmath19 denotes the errors of these parameters , @xmath20 is the mean orbital longitude at @xmath21 .
the above expression is fitted to the pulse arrival times data .
table 2 presents the timing solution of oao 1657 - 415 from our rxte observations .
the pulse frequency derivative obtained from the quadratic trend of the pulse timing analysis is @xmath22 hz s@xmath23 .
the average pulse frequency derivative we deduce from the pulse frequency history of batse archival data is @xmath24 hz s@xmath23 , consistent at the 1@xmath25 level .
the pulse frequency records of oao 1657 - 415 have shown that the source has stochastic spin - up / down trends ( baykal 1997 , bildsten et al .
1997 ) at timescales longer than weeks . in intervals between stochastic changes ,
1657 - 415 shows secular spin - up or down trends with lower values of noise strength ( baykal 1997 ) .
our observations detect the source spinning down , on average .
there are local deviations from quadratic trends which can be interpreted as very short term fluctuations ( see fig .
the torque noise analysis of oao 1657 - 415 from the batse observations ( baykal 1997 ) showed that pulse frequency fluctuations at shorter than @xmath26 days are almost not detectable due to the measuremental noise , however rxte observations are yielding significant fluctuations around the average quadratic trend ( or spin - down rate ) , as shown in fig .
3 . in rxte observations we are able to construct pulses for time intervals as short as a few hundred seconds .
this yields better timing at shorter time scales .
therefore the fluctuations in arrival times less than @xmath26 days are resolved .
the x - ray flux and pulse frequency derivative correlations of oao 1657 - 415 were investigated by using batse archival data base ( baykal 1997 ) .
batse pulse flux at 20 - 60 kev and pulse frequency series have shown no correlation between x - ray flux and pulse frequency derivatives .
a strong correlation between specific angular momentum ( l ) and pulse frequency derivatives was found instead .
these correlations implied that the specific angular momentum is directional , sometimes positive and sometimes negative ( @xmath27 l ) , and that sometimes the flow is radial .
these results suggested the formation of temporary accretion disks in the case of stellar wind accretion and the short term disk reversals are quite possible . in the analysis of batse data ,
the shortest time scales for resolving the significant pulse frequency fluctuations were of the order of 8 days . in the present work , in order to resolve pulse frequencies at shorter time scales , we used high counting statistics of rxte / pca detectors and obtained pulse arrival times at approximately six hours intervals .
for each 4 or 5 pulse arrival times , we fitted a straight line segment and using the slope of this line we estimated a correction to the average pulse frequency ( @xmath28 ) . in this way , we obtained pulse frequency records for each day roughly . in fig .
4 we present the pulse frequency records estimated in this work .
we accumulated the spectral data at 2 - 50 kev , corresponding to observations of pulse arrival times . in each set of observations we fitted a power - law with high energy cut - off with a gaussian emission line
, then we estimate unabsorbed flux ( or intrinsic pulsar flux ) .
5 presents the pulsar flux history during the rxte observations . during the observation , bolometric x - ray flux increased roughly by 70@xmath7 .
the spin - down rate is decreased and a marginal spin - up trend is seen .
this is the first evidence in oao 1657 - 415 to show positive marginal correlation between the x - ray flux and pulse frequency changes .
oao 1657 - 415 , has shown strong spin - up / down torques in its time history which can not be explained by wind accretion ( baykal 1997 ) .
the formation of temporary accretion disks was therefore considered .
if accretion onto the neutron star is from a keplerian disk ( ghosh & lamb 1979 ) , the torque on the neutron star is given by @xmath29 where @xmath30 is the specific angular momentum added by a keplerian disk to the neutron star at the inner disk edge @xmath31 where @xmath32 is the alfven radius , @xmath33 is the neutron star magnetic moment , @xmath34 is a dimensionless function that measures the variation of the accretion torque as estimated by the fastness parameter @xmath35 . here
@xmath36 is the critical fastness parameter where the accretion torque is expected to vanish at @xmath37 depending on the structure of the disk . according to this model
the torque will cause a spin - up if the neutron star is rotating slowly ( @xmath38 ) in the same sense as the circulation in the disk , or spin - down , if it is rotating in the opposite sense . even if the neutron star is rotating in the same sense as the disk flow , the torque will spin - down the neutron star if it is rotating too rapidly ( @xmath39 ) . in such a model
one should see positive correlation between pulse frequency derivative ( @xmath3 ) and moderate mass accretion rate ( @xmath40 ) if the disk is rotating in the same sense as the neutron star ( baykal 1997 ) .
recent observations of accreting neutron stars have shown stochastic spin - up / down trends on time scales from days to a few years ( bildsten et al .
some of the sources switch from spin - up to spin - down states without showing great changes in their mass accretion rates ( bildsten et al .
these unusual behaviors led nelson et al .
( 1997 ) to the possibility of retrograde circulation of accretion disks .
gx 4 + 1 shows correlation between the x - ray flux and the spin - down rate ( chakrabarty et al . 1997 ) which may suggest a retrograde accretion disk . in our rxte observations ,
oao 1657 - 415 has shown marginal correlation with accretion rate and pulse frequency change .
this positive correlation strongly suggesting that the disk formed in the spin - down episode is in prograde direction . from the batse observations ,
it was concluded that the pulse frequency derivatives and x - ray flux were not well correlated and the 8 days was the minimum for the correlation time scale ( baykal 1997 ) .
this rxte observation implies that the time - scale of correlation is short , only a few days . to see the exact nature of correlations between x - ray flux and pulse frequency derivatives ,
an even more extensive broad band x - ray observation should be carried out .
thanks ali alpar and mark finger for critical reading of the manuscript and jean .h .
swank , tod .e .
strohmayer , mike stark , for stimulating discussions and usra for supporting a visit to the gsfc .
@xmath44 @xmath42 confidence intervals are quoted at the 1 @xmath43 level .
+ @xmath45 orbital parameters are taken from bildsten et al . , ( 1997 ) .
p@xmath46=orbital period , a@xmath47 sin i = projected semimajor axis , e = eccentricity , w = longitude of periastron . + | we present new x - ray observations of the high mass x - ray binary ( hmxrb ) pulsar oao 1657 - 415 , obtained during one orbital period ( 10.44 days ) with the rossi x - ray timing explorer ( rxte ) . using the binary orbital parameters , obtained from burst and transient source experiment ( batse ) observations , we resolve the fluctuations in the pulse frequency at time scales on the order of one day for the first time .
recent batse results showed that oao 1657 - 415 has spin - up / down trends in its pulse frequency time series , without any correlation with the x - luminosity at energies @xmath020 kev ( baykal 1997 ) . in the present rxte observations
the source is found to be in an extended phase of spin - down .
we also find a gradual increase in the x - ray luminosity which is correlated with a marginal spin - up episode .
the marginal correlation between the gradual spin - up ( or decrease in spin - down rate ) and increase in x - ray luminosity suggests that the oao 1657 - 415 is observed during a stable accretion episode where the prograde accretion disk is formed . |
the interaction between a quantum system and its environments leads to loss of quantum coherence , or decoherence , in the system .
understanding decoherence is crucial for grasping the boundary between quantum and classical physics .
@xcite it is also essential for testing theories describing quantum measurements .
@xcite from an engineering point of view , the decay of coherence in quantum bit devices ( qubits ) is the most important obstacle for constructing a working quantum computer .
solid state qubits are leading candidates in the projects of designing quantum circuits , where the coherence times of the qubits are required to be sufficiently long to allow for manipulations and transfer of information by logical gates .
the most important source of decoherence in many realizations of solid state qubits are believed to be bistable fluctuators two level systems ( tlss ) , present as tunneling states in the amorphous substrate @xcite used to fabricate the qubit , or in the tunneling junction in superconductor - based devices .
@xcite these tlss are quantum mechanical systems that are , in turn , coupled to their own environments , which are conventionally considered as uncorrelated thermal baths
. usually one does not worry about the fine details of the environment of the tlss , but rather uses simplified models .
the most popular is the bloch - redfield approach , @xcite where the environment is taken into account by introduction of the relaxation and decoherence rates of the tlss .
if the tlss couple more strongly to their own environment than to the qubit , they are usually treated classically .
this means that the dynamical description of the tlss is replaced by a classical dynamics of a fluctuator , which switches randomly between its two metastable states according to a random telegraph process ( rtp ) .
@xcite this approach is often referred to as the spin - fluctuator model .
@xcite in many cases , however , the decoherence of the qubit is determined by only a few fluctuators that are more strongly coupled to the qubit than others .
@xcite in such cases , one might question the validity of the classical model . from a practical point of view
, it is therefore important to know when such a simplified classical description can replace the full quantum mechanical one .
it is also of more fundamental interest in view of the decoherence approach to the quantum - classical transition .
@xcite in this paper , we will develop a simple model allowing to show when a quantum system can in practice be replaced by a classical one , in the sense that interference effects can no longer be observed due to the entanglement with the environment .
however , we believe that this is only a question of a system becoming _ in practice _ classical , i.e. , when we can use a classical model to calculate a physical property of a quantum system . it does not shed any light on the real limitation of quantum mechanics such as the measurement problem , where one discuss deviations from linear quantum mechanics , see ref . for a discussion .
previously , the boundary between quantum and classical regime for the tls has been explored in a model where the qubit is coupled to an impurity state , and an electron can tunnel between this state and an electron reservoir ( metal ) .
@xcite the qubit dephasing rate calculated in the quantum model was found to converge to the classical result in the high - temperature limit . in the study by abel and marquardt ,
@xcite a threshold for strong coupling between the qubit and the tls was defined by the onset of visibility oscillations in the qubit as a function of the ratio between the coupling to the qubit and the reservoir . the threshold for visibility oscillations
was found for higher values of the qubit coupling in the quantum model compared to the classical model , the thresholds finally converge at high @xmath0 , where @xmath1 is the tls - reservoir coupling . thus , both in the decoherence rate and in the visibility oscillations
the classical limit is recovered at high temperature . in this model ,
the temperature plays a dual role : it affects both the energy relaxation rate of the tls , which maps to the switching rate of the rtp , and it affects the dephasing rate of the tls .
the usefulness of separation of the two effects is seen by the fact that it is perfectly possible to consider finite - temperature classical fluctuators by using an asymetric rtp.@xcite this is never obtained in any limit of the model discussed in refs . .
the subsequent considerations are based on the following qualitative picture : the dephasing of the qubit is caused by the generation of entanglement between the qubit and the environment .
if the qubit and the tls are strongly coupled , then they behave as a combined four - level quantum system and the quantum nature of the tls will be important . in such a situation one can not replace it by a classical rtp . on the other hand , if the tls is sufficiently strongly coupled to the environment , it means that the information about its state is continuously transferred to the environment and this prevents any quantum interference to take place . from this
we can guess that the relevant quantity determining whether the tls can be considered either classical or quantum is the ratio of the qubit - tls coupling ( which determines the rate of entanglement generation between the qubit and tls ) and the tls dephasing rate .
the goal of this paper is to study the applicability of the classical model for qubit decoherence due to a tls . in order to achieve this
we study a model where the dephasing rate of the tls can be varied independently of the temperature , so that the classical limit can be taken at any temperature and correspond to the proper assymetric rtp .
we investigate the regime where the tls is coupled weakly to the qubit . by use of a model borrowed from the study of tlss in glasses ,
we compare the pure decoherence rate of the qubit subject to either a quantum tls , in turn coupled to its environment , or a classical fluctuatur , modeled by random telegraph noise .
our model allows us to separate the effects of temperature , coupling to the bath , and decoherence rate of the tls .
we find that the difference in the qubit decoherence rate predicted by the quantum model and the classical one depends on the ratio , @xmath2 , where @xmath3 is the qubit - tls coupling strength and @xmath4 is the decoherence rate of the tls in the pointer basis .
we start by describing the quantum mechanical model for the tls .
the model we use for the tls originates in the study of tunneling states in glasses , i.e. , a particle , or a group of particles that can be approximated by a single configuratiol coordinate in a double - well potential .
@xcite if the particle is charged , it may give rise to a potential on the qubit that depends on its position in the double - well . following phillips , @xcite the hamiltonian for the coupled qubit - tls
is split into the hamiltonians @xmath5 for the qubit , @xmath6 for the tls , @xmath7 for the qubit - tls interaction , @xmath8 for the environment and @xmath9 for the tls - environment interaction : @xmath10 where the pauli matrices @xmath11 , @xmath12 are operators in the hilbert spaces of the qubit and the tls , respectively .
the energy splitting , @xmath13 , and the tunnel amplitude , @xmath14 , can be calculated from the shape of the double - well potential .
@xcite the energy of the qubit depends on the position of the particle in the double - well ( we will in the following refer to the eigenstates of @xmath15 as the position basis ) , and the coupling strength is given by @xmath3 . in this work
, we will assume the simplified case where the qubit does not directly interact with the environment and therefore has no intrinsic dynamics in the absence of the tls .
furthermore , we consider a model where the qubit is subject to pure dephasing @xmath16=0 $ ] , there is no energy relaxation of the qubit in this model and the decoherence of the qubit is therefore insensitive to the qubit energy splitting @xmath17 . when energy relaxation is present , coherent beatings between the qubit and resonant fluctuators are observed .
@xcite in this strong coupling regime , the fluctuator has to be treated as a quantum system .
our present work concentrates solely on non - resonant fluctuators , which are typically modeled classically .
the double - well potential is in general perturbed by electromagnetic and strain fields modifying the asymmetry energy @xmath13 , while perturbations of the barrier height can usually be ignored .
@xcite in our model we therefore assume that the environment couples to the tls in the position basis , i.e. , the eigenbasis of @xmath15 . rather than formally specifying @xmath8 and @xmath9 we consider two kinds of interaction between the tls and the external environment , resonant and non - resonant .
resonant interaction , e.g. , phonons with frequency close to the eigenfrequency of the tls , are responsible for direct transitions between the eigenstates of the tls , @xmath18 and @xmath19 .
we model this interaction by use of the generalized measurement operators defined for a small time step @xmath20 as@xcite @xmath21 here @xmath22 is the identity matrix in the hilbert space of the qubit and the matrices @xmath23 ` measures ' whether the tls is in the ground ( excited ) state , projects the tls onto this state and flips it .
the rates for absorbtion and emission are @xmath24 = \frac{\gamma_1e^{e / t}}{e^{e / t}-1}.\end{aligned}\ ] ] here @xmath25 is the temperature , @xmath26 is the planck distribution , and @xmath27 is the energy splitting of the tls
. the non - resonant interaction does not cause transitions between the eigenstates of the tls .
however , we might assume that in general the state of a phonon interacting with the tls is perturbed by the interaction , and that the perturbation depends on the position of the system in the double - well .
schematically we can write @xmath28 where @xmath29 index the state of the tls in the position basis , @xmath30 is the initial state of the phonon and @xmath31 is the state of the phonon after the interaction , conditioned upon that the tls was initially in the state indexed by @xmath32 .
the interaction , eq .
, results in entanglement between the phonon and the tls , reducing the coherence of the latter .
the rate of decoherence due to non - resonant phonons depends on the overlap element @xmath33 and on the rate of phonons interacting with the system .
we model this interaction by the single parameter @xmath34 , which is responsible for the decay rate of the off - diagonal density matrix elements of the tls in the position basis . , while the resonant phonons are responsible for relaxation parallell to the z-axis , which is the eigenbasis of the tls .
we define the pointer basis by the basis in which the equilibrium density matrix @xmath35 is diagonal .
the rate of decay perpendicular to this axis is denoted by @xmath4 . ] in this model the equilibrium density matrix of the tls will not necessarily lie along the @xmath36-axis of the bloch sphere .
the equilibrium density matrix is determined by the rate @xmath34 of non - resonant phonons responsible for decay perpendicular to the @xmath36-axis on the bloch - sphere and by relaxation along the @xmath37-axis due to resonant phonons in the eigenbasis of the tls , at the rate @xmath38 towards a level determined by @xmath25 .
the situation is illustrated in fig .
[ coord ] .
we define the decoherence rate of the tls , @xmath4 , by the rate at which the off - diagonal density matrix elements decay in the basis where the density matrix is diagonal in equilibrium .
the time evolution in the quantum model is obtained by numerical integration of the von neumann equation for the hamiltonian given by eq .
, with two modifications .
we add a damping term @xmath34 to our differential equation @xmath39\left|\alpha'\right\rangle-\lambda_{\alpha\alpha'}\rho_{\alpha\alpha'},\end{aligned}\ ] ] where @xmath40 is the density matrix of the system composed of the qubit and the tls and @xmath41 which determines the decay of the off - diagonal density matrix elements of the tls in the eigenbasis of @xmath15 .
in addition , the tls absorb and emit phonons at the rates @xmath42 and @xmath43 .
the absorption and emission of phonons is implemented as follows :
for each timestep @xmath20 we make a transform to the eigenbasis of the tls @xmath44 using the rotation matrix @xmath45 -\sin\frac{\theta}{2 } & \cos\frac{\theta}{2 } \end{array } \right ) , \quad \theta\equiv \arctan\left(\frac{\delta_0}{\delta}\right).\ ] ] the density matrix is then updated according to the rates of absorption and emission as @xmath46 before we make the inverse transform @xmath47 , back to the position basis . here
@xmath48 is the density matrix after the ( potential ) interaction with the resonant phonons .
pure dephasing of the qubit by a classical telegraph noise can be described by the interaction hamiltonian @xmath49 where @xmath50 is the position of the fluctuator at time @xmath51 . for details on this model
see , e.g. , ref . and references therein .
the probability for the fluctuator to switch from the state @xmath52 to @xmath53 , and from @xmath53 to @xmath52 in the interval @xmath54 is given by @xmath55 and @xmath56 , respectively .
to describe finite temperature we will consider the situation where the flipping rates @xmath57 and @xmath58 of the fluctuator are in general not identical , but the states are symmetric @xmath59 .
the situation with asymmetric switching rates was previously studied in refs . .
the equilibrium average is given by @xmath60 where @xmath61 and @xmath62 is the probability for the fluctuator to be found in the state @xmath63 , respectively .
the relaxation towards equilibrium is exponential with rate @xmath64 .
the decoherence of the qubit is obtained by averaging over the realizations and initial conditions of the noise process @xmath65 . for a given realization of @xmath65 ,
the schrdinger equation yields a superposition of the eigenstates of the qubit with a contribution to the relative phase @xmath66 .
averaged over the realizations of the stochastic process @xmath65 we obtain the qubit coherence @xmath67 . here
we will use the transfer matrix method developed by joynt _ et al .
_ , @xcite where we obtain directly the ensemble averaged bloch - vector of the qubit .
the state of the qubit - fluctuator system can be stored in the six - dimensional vector @xmath68 where @xmath69 is the bloch vector of the qubit conditioned upon the state @xmath63 of the fluctuator .
the propagator for @xmath70 averaged over the individual realizations of the rtp can be expressed as @xmath71 where @xmath72 while @xmath73 and @xmath74 are generators of the so@xmath75 group and @xmath76 is the pauli matrix .
a direct advantage of this approach is that the qubit state conditioned upon whether the fluctuator is in the state @xmath63 , @xmath77 follows directly from @xmath70 .
in order to compare the decoherence of the qubit subject to either the quantum tls , or the classical telegraph noise , we calculate similar relaxation rates towards the equilibrium level in the two models .
first we choose a set of parameters , @xmath13 , @xmath14 , @xmath38 , @xmath34 and @xmath25 for the quantum model and prepare the tls in the initial state @xmath78 . at this preliminary stage
we are not interested in the qubit and consider the tls and its environment decoupled from the qubit .
we compute numerically the equilibrium occupation probabilities , @xmath79 and @xmath80 , of the tls in the position basis , as well as the relaxation rate @xmath64 .
note that both the equilibrium occupations and the relaxation rate are in general complicated functions of all the parameters in our model . in this work
, we always restrict ourselves to the regime where the tls is overdamped @xmath81 , i.e. , the decoherence rate is sufficiently large such that coherent oscillations are not observed in the tls . parameter is needed .
note also that since the states @xmath82 are not eigenstates of the hamiltonian , the occupation numbers @xmath83 are not given by the boltzmann weights at the bath temperature .
the decoherence rate is expressed through the rates @xmath84 and the equilibrium occupancy @xmath85 with the help of eqs . and .
the qubit decoherence rate is in general a sum over multiple rates . for symmetric telegraph noise and pure dephasing the decay of coherence in the qubit @xmath86
is given by @xcite what is exact meaning of @xmath86 @xmath87,\ ] ] @xmath88 , but in the regime where the coupling to the qubit is weak compared to the damping of the fluctuator , @xmath89 , the long - time behavior of the decoherence is strongly dominated by a single rate , @xmath90 we finally compute the decoherence rate , @xmath91 , of the qubit when it is coupled to the same quantum fluctuator from which we calculated the relaxation rate and equilibrium occupations previously , but this time the initial state of the tls is the thermal equilibrium state .
the decoherence rate of the qubit is calculated by numerical simulation of the coupled qubit - fluctuator density matrix @xmath92 , from which we can find the qubit density matrix by tracing out the degrees of freedom of the quantum fluctuator . from the qubit density matrix ,
@xmath93 $ ] , we find the coherence @xmath94 , where @xmath95 and @xmath96 denote the eigenstates of the qubit .
finally , @xmath94 is fitted to the exponential function @xmath97 .
the relative difference in the decoherence rate of the qubit due to classical telegraph noise and the quantum fluctuator is defined as @xmath98 where @xmath91 and @xmath99 are the decoherence rate of the qubit subject to the quantum fluctuator and to the classical telegraph noise , respectively .
, in the decoherence rate of the qubit subject to either classical telegraph noise , or a quantum fluctuator . in units of the tls
energy splitting @xmath100 the parameters of the quantum fluctuator are @xmath101 and @xmath102 , the coupling to the qubit is @xmath103 .
color coding for @xmath104 is shown on the right .
the relaxation rate to equilibrium along the @xmath15-axis is the same for both the quantum and the classical fluctuator .
contours where the ratio @xmath2 is constant are plotted for comparison ( black lines ) . ]
this quantity is presented in fig .
[ contour ] as a function of the dephasing rate of the fluctuator , @xmath34 , and temperature @xmath25 .
we have restricted ourselves to a parameter range where the tls does not undergo coherent oscillations .
it is evident that the relative difference in the qubit decoherence rate is small for strong decoherence of the tls , and for high temperatures . in this case
we can safely use the simple rtp model rather than the much more complicated quantum model .
superimposed on the contours for @xmath104 we have plotted curves where the ratio @xmath2 is constant .
we find that the difference between the quantum and the classical fluctuator depends to a very good accuracy on the ratio @xmath2 . for the qubit coupled to the quantum tls ( black , dashed ) , and the qubit
subject to the classical spin - fluctuator ( blue , solid ) .
the mutual information is larger when both systems are treated as quantum objects , due to quantum entanglement between the two systems . in this simulation
the parameters , in units of @xmath100 are @xmath103 , @xmath101 , @xmath102 , @xmath105 and @xmath106 . ]
when the qubit is put in contact with the quantum tls , the qubit and the tls will in general entangle due to their coupling .
the mutual information , the information about the state of one of the systems that can be inferred by measuring the other , will for the quantum tls have an entanglement contribution in addition to the classical correlation .
the mutual information for the qubit - quantum tls is defined straightforwadly by the von neumann entropy@xcite @xmath107 where @xmath108 , @xmath108 and @xmath109 are the density matrices of the qubit , the tls and the composite system , respectively .
when we treat the qubit subject to a classical telegraph noise , we introduce quantum states @xmath110 corresponding to the states @xmath63 of the rtp and use the formula @xmath111 here @xmath112 is the probability for the telegraph process to be found in the state @xmath63 , @xmath113 is the density matrix of the qubit conditioned upon that the telegraph process is in the state @xmath63 and @xmath114 . the time evolution of the mutual information for a qubit coupled either to the quantum tls or the classical fluctuator is shown in fig .
the entanglement between the two systems builds up at a rate given by the coupling @xmath3 , but is lost to the environment at a rate given by the decoherence rate of the tls , @xmath4 .
the increased information about the qubit encoded in the quantum tls , compared to the classical fluctuator increases the transfer of entropy to the environment , thus increasing the decoherence rate of the qubit .
this effect might explain the positive @xmath104 found for low values of @xmath25 and @xmath34 .
experimentally , since the composite density matrix @xmath109 is required , the mutual information can only be extracted in the case where one has access to measurement on both the qubit and the fluctuator simultaneously . since
the fluctuator by definition is a system of the environment outside our control , this can not be achieved .
however , the mutual information could potentially be studied in two coupled qubits , where one of the qubits are subject to controlled noise and takes the role of the fluctuator .
qubits subject to engineered noise under the control of the experimentalist has been realized in optically trapped @xmath115be@xmath116 ions , @xcite where also the required quantum gates has already been implemented in a similar systems .
in general , the dynamics of a quantum tls in an environment depends on three parameters ; the relaxation rate @xmath38 , the dephasing rate @xmath34 and the temperature @xmath25 determining the equilibrium occupations . in this paper , we use a model where the processes responsible for pure dephasing couple to the position basis , while the relaxation processes take place in the eigenbasis of the tls .
this model was used in order to study the relevance of the classical rtp model for description of decoherence of a qubit .
if the interaction responsible for pure dephasing processes ( characterized by @xmath34 ) is diagonal in the eigenbasis of the tls , i.e. , @xmath117 , it will not have any effect on the decoherence rate of the qubit , as long as the qubit couples weakly to the tls @xmath118 , and the tls is prepared in the thermal equilibrium state .
the tls will in this case _ always behave as a classical fluctuator _ , and can therefore straightforwardly be modeled by the classical telegraph noise .
in general , the difference in decoherence rate @xmath119 depends on the ratio @xmath120 as well as @xmath2 , where the contours of constant @xmath2 in the @xmath121 versus @xmath34 plot , match those of constant @xmath119 for all values of @xmath120 .
furthermore , we notice that our results do not tell us that it is , in principle , not possible to construct a classical telegraph model providing the same decoherence rate for the qubit as the quantum tls , even in the regime where the deviation @xmath104 between the two models are large according to fig .
[ contour ] .
we show that the decoherence rate of the qubit differ in the two models in the case where the relaxation rates of the classical and quantum fluctuator are identical . to the best of our knowledge ,
there exist no general relationship between the quantum tls model and the classical spin - fluctuator model .
therefore , one should be careful in applying the classical telegraph model unless one expects the decoherence rates of the fluctuators to be much larger than the qubit - fluctuator coupling .
however , in systems such as glasses this inequality is usually expected to hold , and the tls can be treated effectively as a classical fluctuator , @xcite with an exception if the system is subject to an external ac field . @xcite the pointer states of a quantum system are defined as the pure states that are the least affected by environmental decoherence .
@xcite it is generally believed that when the dynamics of the system is dominated by the interaction with the environment , the pointer states are the eigenstates of the interaction hamiltonaian .
@xcite on the other hand , when the system is weakly coupled to the environment , the pointer states are assumed to be the eigenstates of the isolated system .
@xcite our model can be considered to interpolate between the two extremes .
if we define the pointer basis as the basis where the bloch - vector of the system lies along the @xmath36-axis in equilibrium , the decoherence rate @xmath4 of the system is the rate of decay of the off - diagonal elements of the density matrix in this basis . in conclusion ,
we have constructed a model for the quantum tls where we can study its effect on the qubit as a function of both temperature and the decoherence rate of the tls due to its interaction with the environment .
we have compared the decoherence rate of the qubit found in this model , and in the widely used classical telegraph noise model .
we find that the difference in the qubit decoherence rates depends on the ratio @xmath2 between the strength of the qubit - tls coupling and decoherence rate of the tls in the pointer basis . in the limit @xmath122
, the tls behaves essentially classically and the qubit decoherence rate can accurately be predicted by the telegraph noise model . | we investigate the decoherence of a qubit coupled to either a quantum two - level system ( tls ) again coupled to an environment , or a classical fluctuator modeled by random telegraph noise . in order to do this
we construct a model for the quantum tls where we can adjust the temperature of its environment , and the decoherence rate independently .
the model has a well - defined classical limit at any temperature and this corresponds to the appropriate random telegraph process , which is symmetric at high temperatures and becomes asymmetric at low temperatures .
we find that the difference in the qubit decoherence rates predicted by the two models depends on the ratio between the qubit - tls coupling and the decoherence rate in the pointer basis of the tls .
this is then the relevant parameter which determines whether the tls has to be treated quantum mechanically or can be replaced by a classical telegraph process .
we also compare the mutual information between the qubit and the tls in the classical and quantum cases . |
recent studies of various collective processes in dust contaminated plasmas ( dp ) @xcite have been of significant interest in relation with linear and nonlinear waves which are observed in laboratory and space plasmas .
an issue of particular importance is the formation of strongly coupled dp crystals by highly charged dust grains , for instance in the sheath region above a horizontal negatively biased electrode in experiments @xcite .
low - frequency oscillations may occur in these mesoscopic dust grain quasi - lattices , in both longitudinal ( acoustic mode ) @xcite and transverse ( in - plane shear acoustic mode , off - plane optic - like mode ) directions , as theoretically predicted and experimentally observed ( see in ref .
@xcite for a review ) . in this paper
, we focus on the nonlinear description of dust grain displacements in a one - dimensional dust crystal , which is suspended in a levitated horizontal equilibrium position where gravity and electric ( or , possibly magnetic @xcite ) forces balance each other . considering the coupling between the horizontal ( @xmath0 ) and vertical ( off - plane , @xmath1 ) degrees of freedom , and an arbitrary inter - grain interaction potential form @xmath2 ( e.g. debye or else ) and sheath potential @xmath3 ( not necessary parabolic )
, we aim in deriving a set of equations which should serve as a basis for forthcoming studies of the nonlinear behaviour of longitudinal and transverse dust lattice waves ( ldlws , tdlws ) propagating in these crystals .
the relation to recent studies of a similar scope ( here recovered as special cases ) is also discussed .
let us consider a layer of charged dust grains ( mass @xmath4 and charge @xmath5 , both assumed constant for simplicity ) of lattice constant @xmath6 .
the hamiltonian of such a chain is of the form @xmath7 where @xmath8 is the position vector of the @xmath9th grain ; @xmath10 is a binary interaction potential function related to the electrostatic potential @xmath11 around the @xmath12th grain , and latexmath:[$r_{nm } = @xmath9th and @xmath12th grains .
the external potential @xmath14 accounts for the external force fields in which the crystal is embedded ; in specific , @xmath15 takes into account the forces acting on the grains ( and balancing each other at equilibrium , ensuring stability ) in the vertical direction ( i.e. gravity , electric and/or magnetic forces ) ; it may also include the parabolic horizontal confinement potential imposed in experiments for stability @xcite as well as , for completeness , the initial laser excitation triggering the oscillations in experiments . considering the motion of the @xmath9th dust grain in both the _ longitudinal _ ( horizontal , @xmath16 ) and the _ transverse _ ( vertical , off plane , @xmath1 ) directions ( i.e. suppressing the transverse in - plane shear component , @xmath17 ) , so that @xmath18 , we have the two - dimensional ( in @xmath19 ) equation of motion @xmath20 where @xmath21 is the ( interaction ) electrostatic field and @xmath22 accounts for all external forces in the @xmath23 direction ( @xmath24 for @xmath25 ) ; the usual _ ad hoc _ damping term was introduced in the left - hand - side of eq .
( [ eqmotion0 ] ) , involving the damping rate @xmath26 due to dust neutral collisions .
we shall assume a smooth , continuous variation of the ( generally inhomogeneous ) field intensities @xmath27 and/or @xmath28 , as well as the grain charge @xmath5 ( which may vary due to charging processes ) near the equilibrium position @xmath29 .
thus , we may develop @xmath30 @xmath31 and @xmath32 where the prime denotes differentiation with respect to @xmath33 and the subscript ` @xmath34 ' denotes evaluation at @xmath35 , viz .
@xmath36 , @xmath37 and so forth .
accordingly , the electric force @xmath38 and the magnetic force @xmath39 ( where the grain magnetic moment @xmath40 is related to the grain radius @xmath41 and permeability @xmath40 via @xmath42 @xcite ) , which are now expressed as @xmath43 @xmath44 and @xmath45 @xmath46 may be combined to give @xmath47 where we have introduced the phenomenological potential @xmath3 @xmath48 the definitions of @xmath49 ( here , the superscript within parenthesis obviously denotes the order in partial differentiation ; @xmath50 ) are obvious : @xmath51 and so forth .
obviously , @xmath52 .
the ( vertical ) force balance equation @xmath53 , viz .
@xmath54 is satisfied at equilibrium .
assuming small displacements from equilibrium , one may taylor expand the interaction potential energy @xmath2 around the equilibrium inter - grain distance @xmath55 ( between @xmath56th order neighbors , @xmath57 ) , i.e. around @xmath58 and @xmath59 , viz .
@xmath60 where @xmath61 denotes the degree of nonlinearity involved in its contribution : @xmath62 is the linear interaction term , @xmath63 stands for the quadratic potential nonlinearity , and so forth . notice that the inter - grain distance @xmath64^{1/2}$ ] also needs to be expanded , i.e. near @xmath65 and @xmath66 , so that @xmath67 . obviously , @xmath68 and @xmath69 denotes the displacement of the @xmath9th grain from the equilibrium position ( @xmath70 . retaining only nearest - neighbor interactions ( @xmath71 ) , we obtain the coupled equations of motion @xmath72 \nonumber \\ \ , + \ , a_{30 } \ , \biggl [ ( \delta x_{n+1 } - \delta x_{n})^{3 } - ( \delta x_{n } - \delta x_{n-1})^{3 } \biggr ] % \nonumber \\ + \ , a_{02 } \ , \biggl [ ( \delta z_{n+1 } - \delta z_{n})^{2 } - ( \delta z_{n } - \delta z_{n-1})^{2 } \biggr ] \nonumber \\ \ , - a_{12}\ , \biggl [ ( \delta x_{n+1 } - \delta x_{n } ) ( \delta z_{n+1 } - \delta z_{n})^{2 } - ( \delta x_{n } - \delta x_{n-1 } ) ( \delta z_{n } - \delta z_{n-1})^{2 } \biggr ] \ , , \nonumber \\ \label{discrete - eqmotion - x}\end{aligned}\ ] ] and @xmath73 \nonumber \\ \ , + \ , 2 \ , a_{02}\ , \biggl [ ( \delta x_{n+1 } - \delta x_{n } ) ( \delta z_{n+1 } - \delta z_{n } ) - ( \delta x_{n } - \delta x_{n-1 } ) ( \delta z_{n } - \delta z_{n-1 } ) \biggr ] \ , \nonumber \\ - \ ,
a_{12}\ , \biggl [ ( \delta x_{n+1 } - \delta x_{n})^2 ( \delta z_{n+1 } - \delta z_{n } ) - ( \delta x_{n } - \delta x_{n-1})^2 ( \delta z_{n } - \delta z_{n-1 } ) \biggr ] \ , , \label{discrete - eqmotion - z}\end{aligned}\ ] ] where we have defined the longitudinal / transverse oscillation characteristic frequencies @xmath74 ( both assumed to be positive for any given form of interaction potential @xmath75 ) and the quantities @xmath76 \ , , \nonumber \\
a_{30 } = \frac{1}{6 m}\,u''''(r_0 ) \
, , \qquad a_{12 } = - \frac{1}{m r_0 ^ 3}\ , \bigl [ u'(r_0 ) - r_0 u''(r_0 ) + r_0 ^ 2 \ , \frac{1}{2 } \,u'''(r_0 ) \bigr ] \ , , \label{coefficients}\end{aligned}\ ] ] which are related to coupling nonlinearities .
the _ gap frequency _ @xmath77 and the nonlinearity coefficients @xmath78 and @xmath79 are related to the form of the sheath environment ( i.e. the potential @xmath80 ) via @xmath81 obviously , the prime denotes differentiation , viz .
@xmath82 and so on . in the above equations of motion ,
we have distinguished the linear contributions of the first neighbors from the nonlinear ones , i.e. the first line in the right hand
side from the remaining ones , in both equations .
note that all of the coefficients are defined in such a way that they bear _ positive _ values for debye type interactions , i.e. if @xmath83 ( @xmath84 is the effective debye length ) since odd / even derivatives are then negative / positive ; however , the sign of these coefficients is not a priori prescribed for a different interaction potential @xmath2 .
indeed , we insist on expressing all formulae in such a manner that a different interaction law may easily be assumed in a _ `` plug in '' _ manner ; in particular , even though the debye potential @xmath85 is widely accepted in dp crystal models , we think of the modification of @xmath75 when one takes into account a magnetic field @xcite or the ion flow towards the negative electrode @xcite .
nevertheless , we provide the explicit form of the coefficients @xmath86 defined above for a debye potential , for clarity , in the appendix . upon careful inspection of the discrete equations of motion above ,
one notices that the lowest order nonlinearity in the longitudinal motion is due to the intergrain interaction law , while nonlinearity in the vertical motion is primarily induced by the coupling to the horizontal component ( and , to less extent , by interactions ) .
adopting the standard continuum approximation , we may assume that only small displacement variations occur between neighboring sites , i.e. @xmath87 where the ( horizontal ) displacement @xmath88 is now expressed via a continuous function @xmath89 .
the analogous continuous function @xmath90 is defined for @xmath91 .
one may now proceed by inserting this ansatz in the discrete equations of motion ( [ discrete - eqmotion - x ] , [ discrete - eqmotion - z ] ) , and carefully evaluating the contribution of each term .
the calculation , quite tedious yet perfectly straightforward , leads to a set of coupled continuum equations of motion in the form @xmath92 \ , + \ , 3 \ ,
a_{30}\ , r_0 ^ 4 \ , ( u_x)^2 \,u_{xx } \ , , \label{eqmotion - gen - continuum - x}\end{aligned}\ ] ] @xmath93 \ , , \label{eqmotion - gen - continuum - z}\end{aligned}\ ] ] where higher - order nonlinear terms were omitted .
we have defined the characteristic velocities @xmath94 and @xmath95 ; the subscript @xmath96 denotes partial differentiation , so that @xmath97 and @xmath98 .
remember that the _ gap frequency _ @xmath77 and the coefficients @xmath78 and @xmath79 are related to the form of the sheath electric and/or magnetic potential via ( [ k12-def ] ) above , viz .
as a matter of fact , all known older results are based on equations which are readily recovered , as special cases , from eqs . ( [ discrete - eqmotion - x ] ) and ( [ discrete - eqmotion - z ] ) and/or their continuum counterparts ( [ eqmotion - gen - continuum - x ] ) and ( [ eqmotion - gen - continuum - z ] ) . in particular , the coupled eqs .
( 1 ) and ( 2 ) in ref .
@xcite are exactly recovered from ( [ discrete - eqmotion - x ] ) and ( [ discrete - eqmotion - z ] ) , upon neglecting @xmath100 , @xmath101 , @xmath78 and @xmath79 and then evaluating all coefficients for a debye type potential . upon switching off the coupling ( i.e. setting @xmath102 ) ,
( [ eqmotion - gen - continuum - x ] ) above recovers exactly the nonlinear eq .
( 13 ) in ref .
@xcite , which was therein shown to model ( nonlinear ) _ longitudinal _ dust grain motion in terms of ( either korteweg - de vries @xcite or boussinesq type ) solitons ; also see eq .
( 2 ) in @xcite ( treating the formation of asymmetric envelope modulated ldlws ) and eq . ( 2 ) in @xcite ( keep only first - neighbor interactions therein , to compare ) . in a similar manner , considering purely _ transverse _ motion ( i.e. setting @xmath103 ) eqs .
( [ discrete - eqmotion - z ] ) and ( [ eqmotion - gen - continuum - z ] ) herein recover exactly the nonlinear eqs .
( 7 ) and ( 8) in ref .
@xcite , where they were shown to model the amplitude modulation of tdlws which is due to the sheath nonlinearity . finally , needless to say ,
the linear limit recovers exactly the known equations of motion for either purely longitudinal or purely transverse motion ( i.e. considering @xmath104 ) .
an exact treatment of the coupled evolution eqs .
( [ discrete - eqmotion - x ] , [ discrete - eqmotion - z ] ) or , at least , the continuum system ( [ eqmotion - gen - continuum - x ] , [ eqmotion - gen - continuum - z ] ) seems quite a complex task to accomplish . even though eq .
( [ eqmotion - gen - continuum - x ] ) may straightforward be seen as a boussinesq type equation @xcite , which is now modified by the coupling , its transverse counterpart ( [ eqmotion - gen - continuum - z ] ) ( for @xmath105 , say ) substantially differs from any known nonlinear model equation , bearing known exact solutions .
therefore , we shall limit ourselves to reporting this system of evolution equations , for the first time , thus keeping a more thorough investigation ( analytical and/or numerical ) of their fully nonlinear regime for a later report .
in order to gain some insight regarding the influence of the mode coupling on the nonlinear profile of the dust lattice waves , we may consider the effects which come into play when the amplitude of the ldlws and the tdlws which are initially uncoupled in the small amplitude ( linear ) limit is increased to a slightly finite ( i.e. non negligible ) value , thus allowing for a weak coupling between the two modes and a tractable appearance of the signature of the ( weak ) nonlinearity in the dynamics .
the standard way for such an approach is via the introduction of multiple space and time scales , viz . @xmath106 and @xmath107 , where @xmath108 and @xmath109 ( @xmath110 is a smallness parameter ) .
the solutions are expanded as : @xmath111 ( plus an analogous expression for @xmath112 ) .
the technical details of the calculation are described e.g. in @xcite and will be omitted here .
we shall apply this reductive perturbation technique to the system obtained from eqs .
( [ eqmotion - gen - continuum - x ] , [ eqmotion - gen - continuum - z ] ) by keeping only the lowest - order nonlinear terms ( i.e. omitting the last line in both equations ) ; we set @xmath113 and @xmath114 for simplicity . note the inevitable ( and qualitatively expected ) complication of the calculation due to the different dispersion laws in the two modes @xcite .
the first - order ( @xmath115 ) equations are uncoupled and may be solved by assuming @xmath116 @xmath117 $ ] ( complex conjugate ) . upon substitution , we obtain @xmath118 ; the remaining ( 3 ) amplitudes are left arbitrary .
this readily yields the known dispersion relations @xmath119 for the ( acoustic ) ldl and the ( optical - like ) tdl mode respectively .
the 2nd - order ( @xmath120 ) equations contain secular ( 1st - harmonic forcing ) terms , whose elimination imposes a pair of conditions in the form : @xmath121 ( where @xmath122 in the following ) , implying that the amplitudes @xmath123 travel at the ( different ) group velocities @xmath124 .
see that @xmath125 ( the tdlw is a _
backward wave _ ) , as immediately obtained from ( [ dispersion]b ) .
the remaining system is then solved for the @xmath34th and the @xmath126nd harmonic amplitudes ( in @xmath127 ) @xcite ; the solution finally obtained is of the form : @xmath128 + \epsilon^2 \ ,
u_2^{(2 ) } \exp 2 i ( k x - \omega_1 t ) + { \rm{c.c . } } ] + \mathcal o(\epsilon^3 ) \nonumber \\
\delta z_n(t ) & \approx & w(x , t ) \approx \epsilon
[ \psi_2 \exp i ( k x - \omega_2 t ) + { \rm{c.c . } } ] + \epsilon^2 \ , \ { w_2^{(0 ) } + [ w_2^{(2 ) } \exp 2 i ( k x - \omega_2 t ) + { \rm{c.c . } } ] \ } + \mathcal o(\epsilon^3 ) \ , .\end{aligned}\ ] ] we henceforth denote the significant amplitudes @xmath129 , @xmath130 and @xmath131 by @xmath132 , @xmath133 and @xmath134 respectively .
the 2nd order correction amplitudes are @xmath135 where @xmath136 and @xmath137 .
the contributions @xmath138 , @xmath139 and @xmath140 are left arbitrary by the algebra and were thus set to zero .
proceeding to the 3rd - order ( @xmath141 ) equations , the elimination of the secular terms together with zeroth order equations provide three explicit conditions , for @xmath142 .
after some tedious algebra , these take the form @xmath143 where @xmath144 is an integration constant ( to be determined by the boundary conditions ) .
the linear dispersion terms @xmath145 are related to the ( curvature of ) the dispersion relations ( [ dispersion ] ) as @xmath146 ( @xmath147 ) ; the group velocities @xmath148 were defined above @xcite .
the nonlinearity coefficients @xmath149 ( @xmath150 , @xmath147 ) and the ` peculiar ' contributions @xmath151 ( involving cross - terms in @xmath152 ) are too lengthy to report here @xcite . observe that , once @xmath144 is determined , one may cast eqs .
( [ nlse012 ] ) into the form of a ( modified , asymmetric ) system of coupled nonlinear schrdinger equations ( cnlse ) .
note that we have avoided the usual envelope ( galilean ) transformation @xmath153 , since it does not simplify this ( asymmetric , with respect to @xmath154 ) system . finally , let us point out , for rigor , that the results in @xcite and @xcite are exactly recovered , from both ( [ n2l2 ] ) and ( [ nlse012 ] ) , in the appropriate
uncoupled mode limits ( namely , @xmath155 and @xmath156 , respectively , for ldlws and tdlws ) .
despite the obvious analytical complication , the physical mechanism underlying the above results is rather transparent .
there is an energy pumping effect between the zeroth - harmonic longitudinal ( displacement ) mode @xmath132 , first put forward in @xcite ( for ldlws , yet long known in solid state physics @xcite ) and the modulated ( low - frequency ) ldl and ( high - frequency ) tdl mode(s ) .
note the strong misfit ( asymmetry ) between the dispersion laws dominating the coupled modes , despite which regretfully no simplifying assumption may be analytically carried out in this continuum model .
we have put forward a comprehensive nonlinear model for coupled longitudinal - to - transverse displacements in a horizontal dust mono - layer , levitated in a sheath under the influence of gravity and an electric and/or magnetic field .
all of the above results are generic , i.e. valid for any assumed form of the inter - grain interaction potential @xmath2 and the sheath potential @xmath80 , and will hopefully contribute to the elucidation of the grain oscillatory dynamics in dust crystals .
consider the debye potential ( energy ) @xmath157 .
defining the ( positive real ) lattice parameter @xmath158 , one straightforward has @xmath159 @xmath160 where the prime denotes differentiation and @xmath161 is a positive integer .
now , combining with definitions ( [ frequencies ] , [ coefficients ] ) , we have : @xmath162 @xmath163 @xmath164 of course , all known previous definitions of ( some of ) these coefficients ( for nearest neighbour interactions ; see in the references cited in the text ) are exactly recovered .
note , finally , that @xmath165 is of the order of ( or slightly higher than ) unity in experiments ; therefore , all coefficients turn out to be of similar order of magnitude , as one may check numerically .
this work was supported by the _ sfb591 ( sonderforschungsbereich )
universelles verhalten gleichgewichtsferner plasmen : heizung , transport und strukturbildung _
german government programme .
support by the european commission ( brussels ) through the human potential research and training network via the project entitled : `` complex plasmas : the science of laboratory colloidal plasmas and mesospheric charged aerosols '' ( contract no .
hprn - ct-2000 - 00140 ) is also acknowledged .
see that , once damping is taken into account ( via @xmath166 ) , all dispersion - related quantities namely @xmath167 , @xmath148 and @xmath145 ( @xmath168 ) here bear an imaginary part .
all of the corresponding expressions readily obtained from eq .
( [ dispersion ] ) and thus omitted here were exactly recovered by the ( tedious ) algebra within our perturbative scheme ; recall , in passing , that @xmath142 are complex numbers , by definition . | quasi - crystals formed by charged mesoscopic dust grains ( dust lattices ) , observed since hardly a decade ago , are an exciting paradigm of a nonlinear chain . in laboratory discharge experiments ,
these quasi - lattices are formed spontaneously in the sheath region near a negative electrode , usually at a levitated horizontal equilibrium configuration where gravity is balanced by an electric field .
it is long known ( and experimentally confirmed ) that dust - lattices support linear oscillations , in the longitudinal ( acoustic mode ) as well as in the transverse , in plane ( acoustic- ) or off - plane ( optic - like mode ) directions . either due to the ( typically yukawa type ) electrostatic inter - grain interaction forces or to the ( intrinsically nonlinear ) sheath environment
, nonlinearity is expected to play an important role in the dynamics of these lattices .
furthermore , the coupling between the different modes may induce coupled nonlinear modes . despite this evidence ,
the elucidation of the nonlinear mechanisms governing dust crystals is in a rather preliminary stage . in this study
, we derive a set of ( coupled ) discrete equations of motion for longitudinal and transverse ( out - of - plane ) motion in a one dimensional model chain of charged dust grains . in a continuum approximation , i.e. assuming a variation scale which is larger than the lattice constant , one obtains a set of coupled modified boussinesq - like equations .
different nonlinear solutions of the coupled system are discussed , based on localized travelling wave anstze and on coupled equations for the envelopes of co - propagating quasi - linear waves . |
one of the most prominent problems associated with gas - filled proportional chambers is the sparking induced by heavily ionizing particles producing large deposits .
amplified by the avalanche process they could reach a critical charge density , related to the raether s
limit @xcite , and could evolve into a discharge . as a result
, a large fraction of the stored charge defining the amplification field is lost .
the presence of discharges _ limits the high rate operation _ of the detector due to the required power supply recovery time , _ reduces the mean life _ of the detector and _ risks the damage of the readout electronics _ due to the high currents reached by these spark processes . to avoid any damage of the electronics , most of the detectors are using an additional protecting circuit , which interfaces the readout strips or pads with the front - end electronics .
recently new developments are undertaken to further improve the spark protection to operate at higher rates .
a promising approach is employing a resistive material on top of the anode plane @xcite .
when a streamer develops the increasing number of charges is deposited at the high resistivity material .
the anode potential generated by these charges reduces the amplification field , quenching the streamer resulting from field loss .
a required condition is that the anode potential must last for a period of time long enough such that the charges present in the streamer have been completely evacuated , which introduces the need for high resistive values .
the resistive interface limits the charge transfer , protecting the electronics , and avoids the total discharge of the mesh reducing the dead time of the detector .
this new technology needs to be characterized and simulated to exploit its possibilities , using the appropriate resistive and capacitive values depending on the needs of each particular application .
micromegas detectors have evolved recently using this kind of protection .
two different topologies are typically implemented , a fully covering resistive layer and micromegas detectors with resistive strips .
we present the description of these new approaches focusing on recent developments .
then , we will introduce a method to calculate the charge dispersion on the resistive material in each topology . for large micromegas detectors ,
a novel protection scheme has been developed by a group working on hard radiation detectors at cern @xcite . according to this scheme
, we are using resistive strips above the readout electrode to overcome the spark problem .
this technique consists in covering the micromegas readout @xcite with a thin insulating layer where resistive strips are plotted on one favored axis ( see figure [ fig1 ] ) . by adding a drift electrode , a mm detector operates as usually in the proportional avalanche mode inducing signals on the resistive strips plane .
the resistive strip is grounded through an embedded high impedance resistor .
the typical strip resistivity is in the order of 100 m@xmath0/cm , and 200 m@xmath0 for the boundary resistor which drives the strip to ground .
an approach of resistive layer topology is called _ piggyback _ resistive micromegas @xcite where a thin resistive layer is deposited on an insulator .
this element is used as substrate for the micromegas which could be fabricated by using the _ bulk _ technology @xcite .
a schematic view of the new structure is shown in figure [ fig2 ] where we can see the various elements : the micromegas structure with the pillars ; the resistive layer ; the insulating substrate and the readout board .
the thickness of the resistive layer is of the order of 10 @xmath1 m .
a particularity of this new detector technique is the fact that the readout electronics are independent ( and thus exchangeable ) from the micromegas structure .
the elements of the structure are optimized in such a way that the electronic signal is not lost through the resistive layer but is propagated to a separate plane , carrying read - out pads or strips , by capacitive coupling . in order to optimize the induced signal by capacitive coupling the thickness of the insulating ceramic ( t@xmath2 ) must be kept small to satisfy the relation @xmath3 the amplification gap ( t@xmath4 ) is about 100 @xmath1 m and it is achieved between a woven stainless steel mesh connected to the high voltage and a resistive anode .
@xmath5 is the dielectric constant of medium 1 ( gas ) and @xmath6 is the dielectric constant of medium 2 ( insulator ) . because the insulator plays at the same time the role of vessel of the detector the thickness should be kept reasonable ( several hundred @xmath1 m ) . in order to satisfy ( [ permrel ] )
the material should have a dielectric constant as high as possible .
good candidates are ceramic insulators having large dielectric constants ( @xmath710 ) .
the value for the amplification gap is the standard 128@xmath1 m , while the ceramic thickness should be thicker than 300@xmath1 m to assure the leak tightness of the detector .
the resistive layer is made of ruthenium oxide ( ruo@xmath2 ) .
this material is extensively used for coating ceramics at high temperature for the preparation of resistors or integrated circuits , and provides good uniformity and a wide range of resistivities , between 1 - 100m@xmath0/@xmath8 .
a simplified approach of a single resistive strip is used in order to calculate the induced signal at the micromegas usual read - out and the charge diffusion through the resistive strip .
the model here described is based on the equivalent electric circuit introduced by @xcite .
this equivalent circuit describes in general a standard micromegas detector covered by a thin insulating layer where resistive strips are placed facing the standard metallic strips , creating an inter - capacitance and therefore an electric transmission line . in order to calculate the charge diffusion ,
the resistive strip is subdivided in a set of differential elements which final element is grounded through a boundary resistor @xmath9 .
the resistive strip has a linear resistivity @xmath10 and is electrically coupled to the standard read - out , defined by @xmath11 , through the mentioned insulating layer which introduces the strip linear capacitance @xmath12 .
the standard read - out is connected in this model to a low impedance value @xmath13 , and the residual capacitance of the pcb board @xmath14 ( see figure [ resistive_rcmodel ] ) .
the potential on the strip is driven by the following equation @xmath15 the evolution of the potential and/or charge in the resistive strip is bounded by the external circuit elements attached to the strip electrode , @xmath13 and @xmath14 , and the fixed resistive value , @xmath9 , attached to the resistive strip .
the first elements ( @xmath13 and @xmath14 ) allow determination of the strip voltage @xmath11 as a function of the resistive strip potential distribution , by integrating the total equivalent current induced at the capacitive surface @xmath16 along the resistive strip length @xmath17 ,
@xmath18 the resistive value @xmath9 provides the boundary condition on the voltage @xmath19 at the beginning of the strip , and its value is related with the voltage gradient at the resistive strip end , @xmath20 the relation ( [ eq : diffusion_1 ] ) describes the potential propagation at the resistive strip .
however , in order to observe such propagation we need a perturbation which is introduced as a time dependent current source spatially distributed and described by the change on linear charge density @xmath21 along the resistive strip .
this current is introduced inside relation ( [ eq : diffusion_1 ] ) by taking into account the expected voltage increase at the resistive nodes , @xmath22 the introduced charge density pretends to emulate the effect of charge induced by a physical avalanche , which would be the connection between our detector read - out and the physical signal . the solution of the coupled relations ( [ eq1 ] ) and ( [ eq : diffusion_2 ] ) for a given input charge density @xmath21 will provide a full description of the signal propagation within the frame of this model ( a detailed solution can be found at @xcite ) .
this model allows to solve particular scenarios and obtain the transition state of the strip receiving a charge stimulus .
figure [ pulseshapes ] shows the readout signal at the resistive boundary resistor and its shape dependency with a current applied at different strip positions for a total strip length of 10 cm .
this method can be also used to obtain the potential profile in the resistive strip for different rate configurations .
figure [ potdistr ] shows the final state potential distribution resulting from an homogeneous rate distribution .
in this section we show the effect on the gain due to the accumulation of charge over time due to slow charge diffusion in the illuminated region for a resistive layer topology .
the charge diffusion relation for a given charge @xmath23 , which follows a gaussian distribution of width @xmath24 , was provided in @xcite @xmath25 \,\,\,\,\,\,\,\,\,\,\ , h=1/rc\ ] ] where the parameter @xmath26 it is related to the surface resistivity @xmath27 and surface capacitance @xmath28 .
the accumulated charge will be translated into a surface potential given by the capacitance of the ceramic and thus , in a reduction of the amplification field . in order to study such reduction a narrow region around the gaussian distributed charge has been chosen .
the evolution on time of the total charge inside this region , defined by the radius @xmath29 , it is integrated from the previous expression and is described by the following relation @xmath30.\ ] ] we can extrapolate this relation to the case of a continuous current flow by assuming this relation is valid for a differential element of charge . integrating these differential contributions over time will allow us to calculate the accumulated charge at any time , @xmath31 \left ( \frac{dq(v_a)}{dt } \right ) dt \end{array}\ ] ] where we should take into account that these differential charge contributions will be related to the local amplification field , modified by the anode potential @xmath32 at the resistive plane at each time interval .
the integral will be calculated by adding the charge contributions at each time interval n , and describing the gain in terms of the relative mesh versus anode potential , @xmath33 where @xmath34 is the number of electrons produced in each interaction , @xmath35 is the interaction rate , @xmath36 is the gain as a function of the anode potential @xmath32 .
the gain as a function of the mesh potential is obtained experimentally by illuminating the piggyback at low rates ( see figure [ fig10 ] ) , expression that will be used afterwards with the corrected potential given by v@xmath37 .
the charge is integrated considering the effect the accumulated charge will have in the gain at each time step .
the simulation of each time step requires to know previously all the charges at each past time interval in order to apply the diffusion relation to them , this makes the computation of each time interval more expensive and the time required has a non - linear growth .
the simulation of the first half an hour with reasonable accuracy can still be performed within a few hours .
a set of simulations was launched at different scanning rates for different rc values in order to observe the effect on the gain drop .
figure [ fig12 ] shows the final gain value obtained after an exposure of 5400 seconds , several simulations were performed at different rates in order to show the result as a function of the interaction flux .
different resistivity and capacitive values are tested from the range of 10m@xmath38 to 10g@xmath38 , at 1pf / mm@xmath39 . .
] as lowest is the resistivity of the material , the longest is the flat gain region , and higher rates could be achieved without expecting an effect on the gain .
experimental measurements performed with an x - ray generator show a behavior similar to the simulated one ( see figure [ fig13 ] ) . .
we report on two recent representative developments on resistive micromegas technology based on resistive layer and resistive strips topologies .
these techniques profit from the resistive coating by increasing the mean life of the detector , reducing the sparks damage of the electronics and improving the rate capability by reducing the full discharge of the cathode .
we showed two methods which allows us to calculate the charge diffusion on the resistive material and predict the effects on the gain in each topology .
the simulated effects are experimentally observed for the case of piggyback detectors .
the knowledge of the relation between the resistive values and the expected gain drop as a function of the rate will allow us to determine the optimum parameters to be used depending on the final application requirements .
j. wotschak , _ development of micromegas muon chambers for the atlas upgrade _ , _ 2012 jinst 7 c02021_. i. giomataris , _ development and prospects of the new gaseous detector `` micromegas '' _ , _ nucl .
instr . and meth .
a _ , * 419 * ( 1998 ) p.239 .
i. giomataris , ph .
rebourgeard , j.p .
robert , and g. charpak , _ micromegas : a high granularity position sensitive gaseous detector for high particle flux environments _ , _ nucl .
instr . and meth .
a _ , * 376 * ( 1996 ) p.29 .
j. galan , _ signal propagation and spark mitigation in resistive strip read - outs _ , _ 2012 jinst 7 c04009_. m.s .
dixit and a. rankin , _ simulating the charge dispersion phenomena in micro pattern gas detectors with a resistive anode _ , _ nucl .
instr . and meth .
a _ , * 566 - 2 * ( 2006 ) p.281 . | the use of resistive technologies to mpgd detectors is taking advantage for many new applications , including high rate and energetic particle flux scenarios .
the recent use of these technologies in large area detectors makes necessary to understand and characterize the response of this type of detectors in order to optimize or constrain the parameters used in its production , material resistivity , strip width , or layer thickness .
the values to be chosen will depend on the environmental conditions in which the detector will be placed , and the requirements in time resolution and gain , improving the detector performance for each given application .
we present two different methods to calculate the propagation of charge diffusion through different resistive topologies ; one is based on a fem of solving the telegraph equation in our particular strip detector scheme , the other is based on a semi - analytical approach of charge diffusion and is used to determine the charge evolution in a resistive layer .
micromegas , charge diffusion |
the possibility of a perfect lens @xcite whose resolution is not limited by the classical diffraction limit has been subject to intense debate by the scientific community during the past two years .
this perfect lens could be realised by using a slab of material with @xmath4 where @xmath3 is the dielectric constant and the @xmath2 is the magnetic permeability .
veselago had observed@xcite that such a material would have a negative refractive index of @xmath5 ( the negative sign of the square root needs to be chosen by requirements of causality ) , and a slab of such a material would act as a lens in that it would refocus incident rays from a point source on one side into a point on the other side of the slab ( see fig .
1 ) . due to the unavailability of materials with simultaneously negative @xmath3 and @xmath2
, negative refractive index remained an academic curiosity until recently when it became possible to fabricate structured meta - materials that have negative @xmath6 and @xmath2 @xcite .
most of the negative refractive materials(nrm ) , so far , consist of interleaving arrays of thin metallic wires ( that provide negative @xmath3 @xcite ) and metallic split - ring resonators ( that provide negative @xmath2 @xcite ) .
although some initial concerns were expressed @xcite that the observed effects in these experiments were dominated by absorption , the recent experiments of refs.@xcite have confirmed that negative refractive materials are today s reality . and
@xmath7 is refocussed into a point on the other side .
the rays representing propagating waves are bent on to the other side of the normal at the interfaces due to the negative refractive index of the slab .
, width=302 ] it was demonstrated by one of us that the nrm slab acts a lens not only for the propagating waves ( for which the ray analysis of veselago is valid ) but also for the evanescent near - field radiation @xcite .
this phenomenon of perfect lensing becomes possible due to the surface plasmon states @xcite that reside on the surfaces of the nrm slab which restore the amplitudes of the decaying evanescent waves @xcite .
indeed , it has been confirmed by numerical ( fdtd ) simulations that an incident pulse is temporarily trapped at the interfaces for a considerable time @xcite . for a detailed description of the perfect slab lens ,
we refer the reader to ref .
the ` perfectness ' of the perfect lens is limited only by the extent to which the constituent nrm are perfect with the specified material parameters .
absorption in the nrm and deviations of the material parameters from the resonant surface plasmon conditions of the perfect lens causes significant degradation of the subwavelength resolution possible @xcite .
we have suggested some possible measures to ameliorate this degradation of the lens resolution by stratifying the lens medium @xcite and introducing optical gain into the system @xcite .
the image formed by the nrm slab lens is identical to the object and hence there is no magnification in the image .
lenses are mostly used to produce magnified or demagnified images and the lack of any magnification is a great restriction on the slab lens on which most of the attention in the literature has been focussed .
the slab lens is invariant in the transverse directions and conserves the parallel component of the wave - vector . to cause magnification
this tranverse invariance will have to be broken and curved surfaces necessarily have to be involved .
the perfect lens effect is dependent on the near - degeneracy of the surface plasmon resonances to amplify the near - field , and curved surfaces in general have completely different surface plasmon spectrum @xcite .
it was recently pointed out by us that a family of near - field lenses ( in the quasi - static approximation ) in two - dimensions can be generated by a conformal mapping of the slab lens @xcite .
thus a cylinderical annulus with dielectric constant @xmath8 was shown to have a lens - like property of projecting in and out images of charge distributions .
similiarly in ref .
@xcite and @xcite , it was shown how a general method of co - ordinate transformations could be used to map the perfect slab lens solution for the maxwell s equations into a variety of situations including the cylinderical and spherical geometries respectively . in this paper
, we elaborate on the perfect lens solutions in the spherical geometry and show that media with spatially dispersive dielectric constant @xmath9 and magnetic permeability @xmath10 can be used to fabricate a spherical perfect lens that can magnify the near - field images as well . in section-2 of this paper
, we will present these perfect lens solutions of the maxwell s equations for the spherical geometry . in section-3
, we will examine the solutions in the extreme near - field limit or the quasi - static limit which is useful when the lengthscales in the problem are all much smaller than a wavelength .
then the requirements for tm and te polarizations depend only on @xmath11 or @xmath12 respectively .
this is useful at frequencies where we are able to generate structures with only one of @xmath3 or @xmath2 negative .
we will investigate the effects of dissipation in the nrm and point out the connections to the 1-d slab lens solutions .
we will present our concluding remarks in section-4 .
consider a spherically symmetric system shown in fig .
2 consisting of a spherical shell of nrm with the dielectric constant @xmath13 and @xmath14 imbedded in a positive refractive material with @xmath15 and @xmath16 .
first of all we will find the general solutions to the field equations with spatially inhomogeneous material parameters : @xmath17 and @xmath18 images a source located inside the shell into the external region .
the media outside have postive refractive index , but @xmath19 and @xmath20 .
the amplification inside the spherical shell of the otherwise decaying field is schematically shown.,width=302 ] under these circumstances of spherical symmetry , it is sufficient to specify the quantities @xmath21 and @xmath22 which will constitute a full solution to the problem .
let us now look at the tm polarised modes @xmath23 , implying that only the electric fields have a radial component @xmath24 .
operating on eqn .
( 1 ) by @xmath25 , we have @xmath26 , \nonumber \\ & = & \frac{\omega^2}{c^2 } \mu(\mathbf{r } ) \varepsilon(\mathbf{r } ) \mathbf{e } + i \omega \frac{\nabla \mu(\mathbf{r})}{\mu(\mathbf{r } ) } \times \nabla \times \mathbf{e}.\end{aligned}\ ] ] using eqns.(2 ) and ( 3 ) we have @xmath27 = \nabla\varepsilon(\mathbf{r } ) \cdot \mathbf{e } + \varepsilon(\mathbf{r})~ \nabla \cdot \mathbf{e } = 0,\ ] ] and if we assume @xmath28 and @xmath29 , we have @xmath30 we note the following identities for later use : @xmath31 @xmath32 and using eqn . (
6 ) we also note that @xmath33 we now take a dot product of @xmath34 with eqn . ( 4 ) , and use the eqns .
( 6),(7 ) , ( 8) and ( 9 ) to get an equation for @xmath35 as : @xmath36 + \frac{\varepsilon'(r)}{r \varepsilon(r ) } ( re_r ) + \varepsilon(r )
\mu(r ) \frac{\omega^2}{c^2 } ( re_r ) = 0 .\ ] ] this equation is separable and the spherical harmonics are a solution to the angular part . hence the solution is @xmath37 where the radial part @xmath38 satisfies @xmath39 + \frac{\varepsilon'(r)}{r \varepsilon(r ) } u + \varepsilon(r ) \mu(r ) \frac{\omega^2}{c^2 } u = 0.\ ] ] if we choose @xmath40 and @xmath41 , we can have a solution @xmath42 and we get @xmath43 r^{n-2 } + \alpha\beta\omega^2/c^2 r^{p+q+n } = 0,\ ] ] implying @xmath44 and @xmath45.\ ] ] hence the general solution can be written as @xmath46 y_{lm}(\theta,\phi),\ ] ] and @xmath47 \mathbf{x}_{lm}(\theta,\phi),\ ] ] where the vector spherical harmonic @xmath48 a similiar solution can be obtained for the te modes with @xmath49 .
let us note a couple of points about the above perfect lens solutions in the spherical geometry .
first , for @xmath64 , i.e. points outside the image surface the fields appear as if the source were located on the spherical image surface ( @xmath65 ) .
however , this is not true for points @xmath66 within the image surface .
second , given that @xmath67 , we have the perfect lens solutions if and only if @xmath68 which implies that @xmath54 in eqn . ( 13 ) .
although the solutions given by eqn .
( 14 ) occur in any medium with @xmath69 , the perfect lens solutions only occurs for @xmath70 . here
we have written down the solutions for the tm modes .
the solutions for the te modes can be similiarly obtained .
as it has been pointed out in the previous section , the ` power ' solutions are good for any @xmath71 and @xmath72 such that @xmath73 .
however the perfect lens solutions for the maxwell s equations result only for the single case of @xmath74 . in the quasi - static limit of @xmath75 and @xmath76
, we can relax this condition .
in particular , by setting @xmath77 and @xmath78 constant , we can have a perfect lens for the tm modes alone .
similiarly , we can have a perfect lens for the te polarization by having @xmath12 and @xmath79 constant .
this extreme near - field limit is both important and valid for situations when all lengthscales in the problem are much smaller than a wavelength of the radiation .
this becomes useful at frequencies where we can only generate media with either negative @xmath3 and positive @xmath2 , or , negative @xmath2 and positive @xmath3 .
examples are the silver slab lens at optical frequencies @xcite , the imeta - materials ( swiss rolls ) used for mri at radio - frequencies@xcite . particularly at radio- and microwave frequencies
, we currently can practically engineer the required meta - materials with spatially dispersive characteristics at the corresponding length scales .
further it also lifts the restriction that the system has to have a spatially dispersive material parameters even outside the spherical shell of nrm . in this section
we will work in this extreme near field limit .
then it is sufficient to solve the laplace equation and we present lens - like solutions to the laplace equation below .
consider the spherical shell in fig.2 to be filled with a material with @xmath80 with the inner and outer regions filled with constant dielectrics of @xmath81 and @xmath82 respectively .
let @xmath83 everywhere .
now place a charge @xmath84 at the centre of the concentric spheres and a charge @xmath85 at a distance @xmath86 from the centre inside region-1 .
we will consider the z - axis to be along the dipole axis and make use of the azimuthal symmetry here , although it is clear that our results do not depend on any such assumption of azimuthal symmetry .
thus all our charge and their images will now lie along the z - axis .
now we will calculate the potentials in the three regions that satisfies the laplace equation and the continuity conditions at the interfaces .
the potential in region-1(@xmath87 ) can be calculated to be ( using the azimuthal symmetry ) : @xmath88\ ] ] note that the second term in the above expansion arises due to the dipole within the sphere .
it can be shown ( see appendix-1 ) , that the general form of the potential in region-2 ( @xmath89 ) , where the dielectric constant varies as @xmath90 is @xmath91.\ ] ] in region-3(@xmath92 ) , the potential is given by : @xmath93.\ ] ] now we must match the potentials at the interfaces at @xmath94 and @xmath95 ( put @xmath96 ) to determine the @xmath97 and @xmath98 coefficients .
the conditions of continuity of the potential and the normal component of @xmath99 at the interfaces are @xmath100 we determine the coefficients from these conditions to be : @xmath101[\epsilon_1+\epsilon_2(a_1 ) ] - [ \epsilon_2(a_2)+\epsilon_3 ] [ ( l+1)\epsilon_1-l\epsilon_2(a_1 ) ] \frac{a_2^{2l+1}}{a_1^{2l+1}}\ } } { l(l+1)[\epsilon_1+\epsilon_2(a_1)][\epsilon_2(a_2)+\epsilon_3 ] a_2^{2l+1}+ [ l\epsilon_1 - ( l+1 ) \epsilon_2(a_1)][l \epsilon_2(a_2 ) - ( l+1)\epsilon_3 ] a_1^{2l+1}},\\ a_{2l } & = & \frac{(2l+1)[l \epsilon_2(a_2 ) - ( l+1)\epsilon_3]a_0^l a_1^{-1 } } { l(l+1)[\epsilon_1+\epsilon_2(a_1)][\epsilon_2(a_2)+\epsilon_3 ] a_2^{2l+1}+ [ l\epsilon_1 - ( l+1 ) \epsilon_2(a_1)][l \epsilon_2(a_2 ) - ( l+1)\epsilon_3 ] a_1^{2l+1 } } , \\ b_{2l } & = & \frac{(2l+1)(l+1 ) [ \epsilon_2(a_2 ) + \epsilon_3 ] a_2^l a_0^l a_1^{-(l+2 ) } } { l(l+1)[\epsilon_1+\epsilon_2(a_1)][\epsilon_2(a_2)+\epsilon_3 ] a_2^{2l+1}+ [ l\epsilon_1 - ( l+1 ) \epsilon_2(a_1)][l \epsilon_2(a_2 ) - ( l+1)\epsilon_3 ] a_1^{2l+1 } } , \\ b_{3l } & = & \frac{(2l+1)^2 \epsilon_3\epsilon_2(a_2 ) a_0^l a_2^{2(l+1 ) } a_1^{-1 } } { l(l+1)[\epsilon_1+\epsilon_2(a_1)][\epsilon_2(a_2)+\epsilon_3 ] a_2^{2l+1}+ [ l\epsilon_1 - ( l+1 ) \epsilon_2(a_1)][l \epsilon_2(a_2 ) - ( l+1)\epsilon_3 ] a_1^{2l+1}}. \end{aligned}\ ] ] under the perfect lens conditions @xmath102 we have : @xmath103 hence the potential outside the spherical shell for @xmath92 is @xmath104 which is the potential of a dipole with the positive charge at the origin and the negative charge at @xmath105 , where @xmath106 and of strength @xmath107 as @xmath108 .
thus , on one side of the image ( the region @xmath109 ) the fields of a point charge located at @xmath105 are reproduced .
however it should be pointed out that there is no physical charge in the image location and , the fields on the other side of the image ( i.e. in the region @xmath110 ) do not converge to the fields of the object and can not do so in the absence of a charge in the image . further there is no change in the strength of the charge either .
there is a magnification in the image formed by a factor of @xmath111 .
let us point out the similiarities to the planar slab lens . in both cases ,
the electromagnetic field grows in amplitude across the negative medium when the perfect lens conditions are satisfied at the interfaces : as an exponential(@xmath123 $ ] ) in the planar lens and as a power of the radial distance @xmath124 in the spherical lens .
the decaying solution away from the source is absent in the negative medium in both cases .
further , when the perfect lens conditions are matched at both the interfaces , there is no reflected wave in both the planar slab as well as the spherical lens : i.e. the impedance matching is perfect as well .
in addition this mapping preserves the strength of the charge .
the key differences , however , are the different dielectric constants on either sides of the spherical shell of the negative medium .
this is a direct consequence of the spatial @xmath90 dependence of the negative dielectric constant which relates the two positive dielectric constants to be @xmath125 . but
this need not be a particular restriction as we can use the ideas of the asymmetric lens to terminate the different positive media at some radii beyond@xcite .
the net result is that the image can now be magnified ( or demagnified ) when the image of the charge ( source ) is projected out of ( or into ) the spherical shell , which is true in the 2d cylinderical lens as well @xcite . in the case of a planar slab , it was possible to have the perfect lens effect by satisfying the required conditions at any one interface - not necessarily at both interfaces@xcite .
particularly , in the limit of very large parallel wave vectors the lensing is indeed perfect , although the image intensity differed from the source by a constant factor .
similiarly let us now investigate the effects of having the perfect lens conditions in the case of the spherical lens at only one of the interfaces .
let us consider first , the case of projecting out the image of a point source from inside the spherical shell to outside and enable the perfect lens conditions only at the outer interface @xmath126 and have an arbitrary @xmath127 .
now the @xmath97 and @xmath98 coefficients come out to be @xmath128 } { l\varepsilon_1 - ( l+1)\varepsilon_2(a_1 ) } \\ a_{2l } & = & \frac{(2l+1 ) a_0^l}{l\varepsilon_1 - ( l+1)\varepsilon_2(a_1 ) a_1^{2l+2 } } \\ b_{2l } & = & 0 \\ b_{3l } & = & \frac{(2l+1 ) \varepsilon_3
a_0^{l}}{l \varepsilon_1(a_1/a_2)^2 + ( l+1)\varepsilon_3 } \left ( \frac{a_2}{a_1 } \right)^{2l}\end{aligned}\ ] ] only the growing solution within the negative spherical shell remains . the coefficient of the decaying solution ( @xmath129 ) remains strictly zero .
thus , amplification of the decaying field at least is possible in this case as well .
but there is a finite reflectivity in this case .
however , the solution outside for @xmath92 is not the exact image field of the point source as the coefficient @xmath130 has an extra dependence on @xmath131 through the dependence on the dielectric constants .
moreover , the process does not preserve the strength of the charge due to the different dielectric constants involved .
this should be compared to the solution of the planar asymmetric slab lens where , at least in the electrostastic limit , the system behaved as a perfect lens .
in this case , the system behaves as a spherical asymmetric perfect lens only in the limit of large @xmath132 .
the solution outside the spherical shell is the same when we meet the perfect lens condition on the inner interface just as in the case of the planar slab lens .
however , the reflection coefficient is again non - zero , but different to the earlier case . in either case ,
the fields are largest at the interface where one meets the perfect lens conditions or the interface on which the surface plasmons are excited . media with negative real part of the dielectric constant
are absorptive ( as all metals are ) , and hence we can write the dielectric constant @xmath133 ( note that @xmath134 as well for us to be able to write the solution in the following form ) . consider the first case of projecting out the image of a dipole located within the spherical shell where the potential outside the shell is given by equation ( 27 ) and @xmath130 is given by equation ( 33 ) .
when we have a dissipative negative medium and have the perfect lens conditions at the interfaces on the real parts of the dielectric constant alone , @xmath135 and @xmath136 . in parallel with the case of the planar lens
, we note that the denominator of @xmath130 consists of two terms , one containing a power of the ( smaller ) radius @xmath137 and the other containing a power of the ( larger ) radius @xmath138 .
crucially the amplification of the evanescent fields depends on the possibility that the smaller power dominates by making the coefficient of the larger term as close as possible to zero .
the presence of the imaginary part of the dielectric constant would not allow the coefficient to be zero and the image restoration is good only as long as the term containing @xmath137 dominates in the denominator of @xmath130 , i.e. , @xmath139[-(2l+1)\varepsilon_3+i\varepsilon_i(a_2)]a_1^{2l+1}\ ] ] hence a useful estimate of the extent of image resolution can be obtained by noting the multipole @xmath131 for which the two terms in the denominator are approximately equal@xcite .
we obtain for this value : @xmath140 \right\}}{2 \ln \left ( a_2/a_1 \right ) } .\ ] ] higher order multipoles are essentially unresolved in the image . we can similiarly obtain the same criterion by considering the second case of tranferring the image of a charge located outside the spherical shell into the inner region .
again , we can consider the effects of deviating from the perfect lens conditions on the real part of the dielectric constant as well and obtain a similiar limit for the image resolution due to those deviations .
in conclusion , we have presented a spherical perfect lens which enables magnification of the near - field images .
the perfect lens solution requires media with @xmath0 and @xmath141 and the conditions @xmath142 and @xmath143 at the interfaces of the spherical shell of the nrm .
we have shown that in the quasi - static limit of small frequencies ( @xmath144 ) and high - order multipoles @xmath145 , this condition can be relaxed and the two polarizations ( te and tm modes ) decouple .
thus a shell with negative dielectric constant @xmath146 with @xmath147 can act as a near - field lens for the tm polarization while @xmath148 with @xmath149 acts as a near - field lens for the te modes .
we have shown that dissipation in the lens material , however , prevents good resolution of higher order multipoles .
thus while the near - field lenses work best for the higher order multipoles , dissipation cuts - off the higher order multipoles .
further the spherical lens works in the asymmetric mode only in the limit of high order multipoles .
thus one has to find an intermediate regime where dissipation does not wipe out the near - field image information and yet the meta - materials work .
this is the design challenge involving these near - field lenses .
we have to solve the maxwell s equations in material media @xmath150 using @xmath151 , where @xmath152 is the electrostatic potential we have : @xmath153 if @xmath154 has only a radial dependence ( as in our case @xmath155 ) , then @xmath156 and we can separate the solution as @xmath157 where the @xmath158 is the spherical harmonic and the radial part @xmath38 satisfies : @xmath159 = 0\ ] ] to have a solution as a single power of @xmath160 , the only choices possible for the dielectric constant are either @xmath161 , a constant the usual case , or @xmath162 . in the latter case
the solution is @xmath163 or @xmath164 .
the full solution can then be written as @xmath165 y_{lm}(\theta,\phi).\ ] ] 99 j.b .
pendry , phys .
* 85 * , 3966 ( 2000 ) .
veselago , usp .
nauk * 92 * , 517 ( 1964 ) [ sov .
uspekhi , * 10 * , 509 ( 1968 ) . ]
smith , w.j .
padilla , d.c .
vier , s.c .
nemat - nasser and s. schultz , phys .
lett . , * 84 * , 4184 ( 2000 ) .
shelby , d.r .
smith and s. schultz , _ science , _ * 292 * , 77 ( 2001 ) .
a. grbic and g.v .
eleftheriades , j. appl . phys .
* 92 * , 5930 ( 2002 ) .
pendry , a.j .
holden , w.j .
stewart and i. youngs , phys .
, * 76 * , 4773 ( 1996 ) ; j.b .
pendry , a.j .
holden , d.j .
robbins and w.j .
steward , j. phys . : condens . matter , * 10 * , 4785 ( 1998 ) .
pendry , a.j .
holden , d.j .
robbins and w.j .
steward , ieee trans .
mtt , * 47 * , 2075 ( 1999 ) .
n. garcia and m. nieto - vesperinas , opt
* 27 * , 885 ( 2002 ) .
parazzoli , r.b .
greegor , k. li , b.e.c .
kotenbah and m. tanielan , phys .
lett . * 90 * , 107401 ( 2003 ) .
houck , j.b .
brock , i.l .
chuang , phys .
lett . * 90 * , 137401 ( 2003 ) .
a. grbic and g.v .
eleftheriades , appl .
. lett . * 82 * , 1815 ( 2003 ) .
parimi , w.t .
lu , p. vodo , j. sokoloff and s. sridhar , cond - mat/0306109 .
h. raether , _ surface plasmons _ , ( springer - verlag , berlin , 1988 ) d.r .
smith , d. schurig , m. rosenbluth , s. schultz , s.a .
ramakrishna and j.b .
pendry , appl .
82 * , 1506 ( 2003 ) .
haldane , cond - mat/0206420 .
g. gomez - santos , phys .
lett . , * 90 * , 077401 ( 2003 ) .
s.a . ramakrishna , j.b .
pendry , d.r .
smith , d. schurig and s. schultz , j. mod .
optics , * 49*,1747 ( 2002 ) x.s . rao and c.n .
ong , phys .
b * 68 * , 113103 ( 2003 ) .
s. foteinopolou , e.n . economou and c.m .
soukoulis , phys .
lett . * 90 * 107402 ( 2003 ) .
pendry and s.a .
ramakrishna , j. phys . : condens .
matter * 14 * , 8463 ( 2002 ) .
shen and p. platzmann , appl .
. lett . * 90 * , 3286 ( 2002 ) z. ye , phys .
b * 67 * , 193106 ( 2003 ) .
n. fang and x. zhang , appl .
. lett . * 82 * , 161 ( 2003 ) .
ramakrishna , j.b .
pendry , m.c.k .
wiltshire and w.j .
stewart , j. mod .
optics * 50 * 1419 ( 2003 ) .
s. a. ramakrishna and j.b .
pendry , phys .
b. * 67 * , 201101(r ) ( 2003 ) .
klimov , opt .
* 211 * , 183 ( 2002 ) .
pendry , opt .
express * 11 * , 755 ( 2003 ) .
pendry and s.a .
ramakrishna , j. phys . : condens .
matter * 15 * , 6345 ( 2003 ) m.c.k .
wiltshire , j.b .
pendry , i.r .
young , d.j .
larkman , d.j .
gilderdale and j.v .
hajnal , science * 291 * , 848 ( 2001 ) ; opt .
express , * 11 * , 709 ( 2003 ) . | it has been recently proved that a slab of negative refractive index material acts as a perfect lens in that it makes accessible the sub - wavelength image information contained in the evanescent modes of a source . here
we elaborate on perfect lens solutions to spherical shells of negative refractive material where magnification of the near - field images becomes possible .
the negative refractive materials then need to be spatially dispersive with @xmath0 and @xmath1 .
we concentrate on lens - like solutions for the extreme near - field limit .
then the conditions for the tm and te polarized modes become independent of @xmath2 and @xmath3 respectively . |
pressure is one of the fundamental thermodynamic variables . the calculation of pressures in fluid systems using computer simulations is generally considered to be a routine task .
however , difficulties arise in the presence of long - range interactions . here
, we investigate the calculation of pressures in computer simulations of charged and polar systems , where the long - range coulomb interactions are commonly treated with ewald lattice summation@xcite or physically equivalent methods like particle - mesh ewald,@xcite kubic - harmonic expansions,@xcite or lekner sums.@xcite a mechanistic definition of the pressure leads to the standard virial expression .
a thermodynamic definition of the pressure is based on the volume dependence of the helmholtz free energy .
when the coulomb interactions are resummed by using , e.g. , the ewald method , the resulting effective pair interactions depend explicitly on the volume .
in addition , self interactions are present that also depend on the volume . as a consequence ,
the virial and thermodynamic pressures are not identical for finite coulomb systems , even though the two pressures are expected to converge in the thermodynamic limit .
the paper is organized as follows : in section [ sec : theory ] , we derive expressions for the virial and thermodynamic pressures . in section [ sec : presscoul ] , we study the pressure in systems with long - range coulomb interactions . for the thermodynamic pressure , we derive a simple formula that can be readily implemented in standard molecular dynamics or monte carlo codes . in sections [ sec : sim ] and [ sec : spce ] , we study the system size dependence of the virial and thermodynamic pressures for the extended simple point charge ( spc / e ) water model@xcite under standard conditions .
the pressure @xmath0 can be calculated from a mechanistic prescription equating the exterior and interior forces on the container .
this leads to the virial expression for the pressure in an atomic system,@xcite @xmath1 where @xmath2 is the number density of particles ; @xmath3 is boltzmann s constant ; @xmath4 is the temperature ( @xmath5 ) ; and @xmath6 is the volume .
the sum extends over the scalar product betwen particle positions @xmath7 and forces @xmath8 exerted on particle @xmath9 due to other particles in the system .
@xmath10 denotes a canonical average . for computer simulations under periodic boundary conditions with pair forces , it is convenient to rewrite the virial equation in a form that makes the translational invariance explicit : @xmath11 where @xmath12 and @xmath13 is the pair force exerted on particle @xmath9 by particle @xmath14 , derived from a pair potential @xmath15 ; and the sum is over all pairs of particles in the system . for a system of rigid polyatomic molecules
@xmath16 with interaction sites @xmath17 and @xmath18 , one obtains an analogous formula when the forces @xmath19 between molecular sites are projected onto a vector @xmath20 between the `` centers '' of the two molecules ( e.g. , the center of mass ) , @xmath21 here , @xmath22 is the net force between two rigid molecules , summed over molecular sites @xmath17 and @xmath18 .
otherwise , the constraint forces maintaining the rigidity of the molecules have to be included explicitly in eq .
( [ eq : virial_pair_mol ] ) .
the thermodynamic expression for the pressure is derived from the relation between the pressure @xmath23 , the helmholtz free energy @xmath24 , and the volume @xmath6 , @xmath25 the statistical - mechanical relation between the free energy @xmath24 and the partition function @xmath26 for @xmath27 identical classical particles in a canonical ensemble is : @xmath28 where @xmath29 is planck s constant ; @xmath30 is the hamiltonian ; and the integration extends over the positions @xmath31 and momenta @xmath32 of all particles . in taking the volume derivative eq .
( [ eq : p_therm ] ) , the kinetic energy @xmath33 is independent of the volume .
transforming the positional coordinates into dimensionless form , @xmath34 , and pulling out a factor @xmath35 from the integral , leads to the ideal gas term @xmath36 for the pressure .
the non - ideal contributions are contained in the volume dependence of the potential energy @xmath37 , @xmath38 typically , @xmath37 does not depend explicitly on the volume .
the volume dependence of @xmath37 then arises from the volume scaling of the particle positions . in the absence of an explicit volume dependence
, we can express @xmath39 as @xmath40 with @xmath41 . by using @xmath42 and combining eqs .
( [ eq : pcanon ] ) and ( [ eq : dudv ] ) , we find the corresponding thermodynamic pressure to be equivalent to the mechanistic pressure eq .
( [ eq : virial ] ) .
the identity between the virial and thermodynamic pressures , eqs .
( [ eq : virial ] ) and ( [ eq : pcanon ] ) , does not hold if the potential depends explicitly on the system volume .
such an explicit volume dependence arises in computer simulations of charged and polar substances , when the long - range coulomb interactions are resummed , e.g. , by using the ewald method.@xcite we first split the total potential energy @xmath37 into a short - range part @xmath43 and a long - range coulomb part @xmath44 , @xmath45 the pressure @xmath46 corresponding to @xmath43 contains the ideal - gas term and the contributions from short - range pair interactions , @xmath47 where the short - range forces @xmath48 are those derived from the short - range part @xmath43 of the potential energy .
note that the virial and thermodynamic expressions for @xmath46 are equivalent , and therefore the subscripts `` _ v _ '' or `` _ t _ '' have been omitted in eq .
( [ eq : pshort ] ) .
next , we consider the pressure arising from the potential energy @xmath44 of long - range coulomb interactions . in ewald lattice summation ,
the charges in a periodically replicated simulation box interact with an effective potential . that potential is obtained from a summation over all periodic images .
in addition , a self interaction arises from interactions with a particle s own images .
this leads to a coulomb energy @xmath44 for a system of partial charges @xmath49 at positions @xmath50 : @xmath51\nonumber\\ & & + \frac{1}{2 } \sum_{i } \sum_{\alpha } { q_{i\alpha}}^2 \lim_{{\bf r}\rightarrow 0 } \left[\varphi({\bf r})-\frac{1}{|{\bf r}|}\right]~.\end{aligned}\ ] ] the first sum is the intermolecular contribution ; the second and third sums are the intramolecular contributions , with the self interactions contained in the third sum .
@xmath52 is the effective , resummed coulomb potential , with a fourier representation:@xcite @xmath53 where the @xmath54 sum extends over the reciprocal lattice corresponding to the lattice vectors @xmath55 of periodically replicated simulation boxes . in a cubic lattice of length @xmath56
, we have @xmath57 , and @xmath58 , where @xmath9 , @xmath14 , and @xmath59 are integers .
it is numerically convenient to partly transform @xmath52 into real space , leading to its ewald lattice sum representation , @xmath60 @xmath61 is a convergence parameter chosen to accelerate numerical convergence .
the value of @xmath52 is independent of @xmath61,@xcite @xmath62 the self - interactions in @xmath44 are given by the interactions of a unit point charge with its periodic images , subtracting the bare self interaction , @xmath63 , with the appropriate limit taken for @xmath64 .
for a given box shape , @xmath52 scales with the box volume @xmath6 as [ eq : phiscale ] @xmath65 where star superscripts denote volume - independent quantities .
this follows from eq .
( [ eq : phi_four ] ) with volume scaling @xmath66 and @xmath67 .
the same scaling is true trivially for the direct @xmath68 interactions . for an ionic system of point charges without bond constraints ,
( [ eq : pcanon ] ) and ( [ eq : phiscale ] ) immediately lead to an expression for the thermodynamic pressure in terms of the coulomb energy @xmath44 , @xmath69 equation ( [ eq : pionic ] ) gives the well - known relation between the pressure and energy of an ionic system , for which the coulomb energy is a homogeneous function of degree @xmath70 in the coordinates.@xcite for a system of rigid molecules , we find the following volume scaling : [ eq : molscal ] @xmath71~,\\ \frac{\partial}{\partial v } \frac{1}{|{\bf r}_{i\alpha i\beta}| } & = & -\frac{1}{3v}\left[\frac{1}{|{\bf r}_{i\alpha i\beta}| } + \frac{\partial}{\partial { \bf r}_{i\alpha i\beta } } \frac{1}{|{\bf r}_{i\alpha i\beta}|}\cdot { \bf d}_{i\alpha i\beta}\right ] \equiv 0~ , \label{eq : drv } \end{aligned}\ ] ] where @xmath20 is the distance vector between two molecule centers ; @xmath72 is the vector from the center to site @xmath17 ; and @xmath73 .
equation ( [ eq : drv ] ) follows from the volume independence of the intramolecular distance vector @xmath74 . combining eqs .
( [ eq : uew ] ) and ( [ eq : molscal ] ) , we find for the volume derivative of the coulomb energy : @xmath75 \cdot { \bf d}_{i\alpha j\beta}\right . \nonumber\\&&\left .
+ \sum_{i } \sum_{\alpha,\beta \atop \alpha<\beta } \left[\frac{\partial}{\partial { \bf r}_{i\alpha i\beta } } q_{i\alpha } q_{i\beta } \left(\varphi({\bf r}_{i\alpha i\beta})-\frac{1}{|{\bf r}_{i\alpha i\beta}|}\right)\right ] \cdot { \bf d}_{i\alpha i\beta } \right\}~.\end{aligned}\ ] ] we can simplify @xmath76 further by expressing it in terms of the intermolecular forces @xmath77 exerted by site @xmath78 onto site @xmath79 , @xmath80 and the intramolecular forces @xmath81 , @xmath82 = - { \bf f}^{\rm ( intra)}_{i\beta i\alpha}~.\end{aligned}\ ] ] this leads to @xmath83 the sums over pairs of sites @xmath84 and @xmath85 can be rewritten as a single sum over all sites .
this is possible because the distances @xmath86 are intramolecular and are continuous when a particle crosses the box boundary ( i.e. , @xmath87 does not change when the periodic images of the particles @xmath9 or @xmath14 are used ) .
this results in @xmath88~,\end{aligned}\ ] ] where @xmath89 is the net intermolecular force on site @xmath84 , @xmath90 and @xmath91 is the net intramolecular force on site @xmath84 , @xmath92 it is advantageous to add the intra and intermolecular forces because in common ewald - sum implementations the fourier term already contains the sum of both inter and intramolecular contributions which are thus not easily separated .
we define the net coulomb force @xmath93 on site @xmath84 as the sum of the inter and intramolecular forces , @xmath94 we then find for the thermodynamic pressure of a system of rigid molecules : @xmath95 thus the presence of intramolecular constraints in rigid polyatomic molecules resulted in a force term to be subtracted from the pressure of the purely ionic system , eq .
( [ eq : pionic ] ) .
note that the forces @xmath96 in eq .
( [ eq : pmol ] ) are derived from the coulomb energy @xmath44 alone .
additional ideal - gas and short - range contributions to the pressure are reflected in @xmath46 .
the infinite ewald lattice is implicitly embedded in a conducting medium with dielectric constant @xmath97 , corresponding to `` tin - foil '' boundary conditions .
this is the appropriate choice for a conducting medium .
however , for a polar substance it can be advantageous to use a reaction - field dielectric constant @xmath98 similar to that of the bulk medium.@xcite the resulting correction to the coulomb energy @xmath99
is@xcite @xmath100 @xmath101 is the instantaneous dipole moment of the simulation volume arising from the dipole moments @xmath102 of individual molecules , @xmath103 for rigid molecules the @xmath102 do not change with volume .
the reaction - field correction , eq .
( [ eq : eps_rf ] ) , thus scales as @xmath104 , @xmath105 the forces derived from the reaction - field correction are @xmath106 by using eq .
( [ eq : m ] ) , we can express the sum of reaction - field forces projected onto the intramolecular distance vectors in terms of the reaction - field energy @xmath107 , @xmath108 accordingly , the volume derivative of the reaction - field energy @xmath109 can be written as @xmath110~.\end{aligned}\ ] ] the correction eq .
( [ eq : eps_rf ] ) for a finite reaction - field dielectric constant @xmath98 then leads to an expression for the thermodynamic pressure analogous to eq .
( [ eq : pmol ] ) , @xmath111~.\end{aligned}\ ] ] here , the forces @xmath112 are derived from the coulomb energy @xmath113 @xmath114 and contain the reaction field contribution @xmath115 defined in eq .
( [ eq : dumdr ] ) .
we emphasize the simplicity of the pressure expressions eqs .
( [ eq : pmol ] ) and ( [ eq : pmol_m ] ) for systems of rigid molecules .
the coulomb energy contribution @xmath116 is analogous to that of the corresponding ionic system , eq .
( [ eq : pionic ] ) , corrected for the presence of constraint forces . a more or less equivalent expression for the pressure in coulombic systems treated with ewald summation
was derived before by smith,@xcite and similarly by boulougouris _
et al._@xcite .
however , in those derivations the volume derivative was carried out explicitly for the ewald energy . also , the derivations start from an approximate expression for the ewald energy that does not include the full real - space lattice sum and self terms
. therefore , the derivations did not arrive at a closed expression and the simplicity of the results given here was masked .
expressions for the pressure tensor @xmath117 for ewald summation were derived previously by nos and klein,@xcite and heyes,@xcite as discussed by alejandre _
et al._,@xcite as well as by others.@xcite however , the tensor character does not lend itself easily to a compact notation for the bulk pressure @xmath118 in a homogeneous system .
equations ( [ eq : pionic ] ) , ( [ eq : pmol ] ) , and ( [ eq : pmol_m ] ) have the advantage of being independent of the specific method used to evaluate the energies and forces .
all that is needed is the total coulomb energy and forces at all sites that are consistent with that energy .
this is what molecular dynamics codes will normally produce at no additional cost .
the coulomb interactions can then be evaluated by using conventional ewald sums,@xcite particle - mesh ewald,@xcite kubic - harmonic expansion@xcite , or lekner sums.@xcite for approximate coulomb energy calculations such as reaction - field@xcite or generalized reaction - field methods,@xcite eqs .
( [ eq : pionic ] ) , ( [ eq : pmol ] ) , and ( [ eq : pmol_m ] ) suggest an evaluation of the pressure that is formally consistent with that of ewald sums and physically equivalent methods .
to investigate the quantitative differences between the virial and thermodynamic pressures , we study a model of water at standard conditions ( 298 k temperature , 997.07 kg m@xmath119 mass density corresponding to a number density of @xmath120 nm@xmath119 ) .
we use the spc / e model of water,@xcite formed by a lennard - jones center on the oxygens , @xmath121 where @xmath122 kj nm@xmath123/mol and @xmath124 kj nm@xmath125/mol .
in addition , the spc / e model carries three partial charges .
the hydrogen and oxygen sites carry charges @xmath126 and @xmath127 , respectively , where @xmath128 is the elementary charge .
the oxygen - hydrogen bond length is 0.1 nm , the hydrogen - oxygen - hydrogen bond angle is @xmath129 deg .
we use metropolis monte carlo simulations for the canonical sampling,@xcite where the translational and rotational move widths are chosen to give an acceptance rate of about 40 per cent .
ewald summation is used for the electrostatic interactions , with @xmath130 where @xmath56 is the length of the cubic box .
a spherical cutoff of @xmath131 is used for the real space interactions ( charge and lennard - jones ) .
the real - space potentials are shifted by a constant , such that they are zero at the cutoff .
the fourier space sum is truncated at @xmath132 , leading to @xmath133 @xmath54 vectors being considered .
a reaction - field dielectric constant of @xmath134 has been used in all simulations .
standard finite - size corrections were applied to the lennard - jones contributions to pressure and potential energy.@xcite system sizes of @xmath135 , 32 , 64 , 128 , 256 , and 512 water molecules are studied . starting from random configurations
, these systems have been equilibrated for at least 250000 monte carlo passes .
( one pass corresponds to one attempted move for each of the @xmath27 particles . ) in the production runs , the energy as well as the virial and thermodynamic pressures are calculated every tenth pass .
table [ tab : p ] contains the simulation characteristics , as well as results for the virial and thermodynamic pressures .
the thermodynamic pressure is calculated using eq .
( [ eq : pmol_m ] ) .
the virial pressure is calculated using eq .
( [ eq : virial_pair_mol ] ) , where the pair forces are derived from the total potential energy @xmath136 . also included in table [ tab : p ] are results for the average potential energy per particle .
( to compare with the experimental energy , one has to add a polarization correction of 5.22 kj / mol.@xcite ) errors are obtained from a block analysis,@xcite plotting calculated standard deviations of the mean as a function of the number of blocks used .
the reported error is then the plateau value reached in the limit of long blocks with typically more than about 10000 monte carlo passes .
the system size dependence of the virial and thermodynamic pressure is shown in fig .
[ fig : p ] . from table [
tab : p ] and fig .
[ fig : p ] , we find that @xmath137 and @xmath23 converge to the same value of about @xmath138 mpa ( 1 mpa = 10 bar ) for large system sizes , with a statistical error of about 2 mpa .
this convergence is expected as the thermodynamic and virial pressure should be identical in the thermodynamic limit . however , the thermodynamic pressure exhibits a considerably weaker system size dependence than the virial pressure .
the thermodynamic pressure for as few as 64 spc / e water molecules is in agreement with large system sizes .
the virial pressure , on the other hand , scales as roughly @xmath139 for small to intermediate system sizes , with its value off by about one order of magnitude for @xmath140 .
we emphasize that for typical system sizes of @xmath141 , the virial and thermodynamic pressures are identical within statistical errors for spc / e water under standard conditions .
figure [ fig : gxx ] shows the radial distribution functions of water oxygens and hydrogens , which were calculated also in the corners of the cubic simulation box with appropriate weights .
we find that the the radial distribution functions for @xmath142 water molecules are practically indistinguishable , whereas the @xmath135 and @xmath143 simulations are somewhat more structured beyond the first peaks .
these slight structural differences could explain the deviations of the thermodynamic pressure for those small system sizes .
we caution that these are results for the specific thermodynamic state ( room temperature and standard density ) studied here , and we expect more pronounced finite - size effects , e.g. , for low densities . in constant pressure
simulations,@xcite the box volume is rescaled according to the `` instantaneous pressure '' obtained from individual configurations by omitting the canonical average @xmath144 in the pressure formulas above .
it is therefore important that not only the average but also the instantaneous pressure driving the volume fluctuations be correct . as measures of discrepancy between the virial and thermodynamic pressures @xmath137 and
@xmath23 , we use the correlation coefficient @xmath145 and the average absolute deviation minus the deviation of the averages , @xmath146 , [ eq : deviation ] @xmath147 where instantaneous pressures @xmath23 and @xmath137 are used .
results for @xmath145 and @xmath146 are listed in table [ tab : p ] .
the cross - correlation coefficient @xmath145 indicates strong correlation , with @xmath145 values between 0.90 and 0.997 for @xmath135 to @xmath148 .
however , the average absolute deviation @xmath146 between the two pressures is significant even for systems of 512 water molecules , scaling approximately as @xmath149 .
therefore , in constant pressure simulations , the use of the thermodynamic pressure appears advantageous . in an earlier study of pressure effects on the stability of hydrophobic aggregates in water,@xcite we determined the thermodynamic pressure of spc water@xcite as a function of density using eq .
( [ eq : pmol_m ] ) .
for the temperature and density studied here ( @xmath150 k , @xmath120 nm@xmath119 ) , we found a pressure of about @xmath151 mpa for spc water . from the density dependence of the pressure
, we determined a compressibility factor @xmath152 for spc water , where @xmath153 is the isothermal compressibility .
that compressibility factor is in excellent agreement with the experimental value of 0.062 .
we have derived a simple , compact expression for the coulomb contribution to the thermodynamic pressure in a system treated with ewald lattice summation . for a system of point ions ,
we recover the well - known relation between the pressure and potential energy .
we then derive an expression for the pressure in a system of rigid molecules carrying point charges .
the pressure in such a system can be calculated from the total energy and the forces at each site alone .
this makes the implementation of that pressure formula trivial , because both energy and forces are produced routinely in molecular dynamics codes .
moreover , these formulas are entirely independent of the particular method used to resum the coulomb interactions .
ewald summation , particle - mesh ewald,@xcite kubic - harmonic expansions,@xcite and lekner sums@xcite can be used readily . for approximate reaction - field methods,@xcite expressions for the pressure
are suggested by analogy .
we have compared the thermodynamic pressure , obtained from the volume dependence of the helmholtz free energy , with the mechanistic virial pressure .
we find that for rigid spc / e water at standard conditions , the two pressures are approximately equal ( within errors of about 2 mpa ) for systems larger than @xmath154 water molecules . for smaller systems ,
the virial pressure exhibits a pronounced system - size dependence , whereas the thermodynamic pressure can be calculated accurately by using as few as 64 spc / e water molecules .
h. j. c. berendsen , j. p. m. postma , w. f. van gunsteren , and j. hermans , in _ intermolecular forces : proceedings of the 14th jerusalem symposium on quantum chemistry and biochemistry _ , edited by b. pullman ( reidel , dordrecht , holland , 1981 ) , pp . 331342 .
.characteristics and results of the monte carlo simulations of spc / e water .
statistical errors are one estimated standard deviation of the mean . also included are the cross - correlation coefficient @xmath145 and the absolute deviation @xmath146 , as defined in eq .
( [ eq : deviation ] ) . [ cols=">,>,>,>,>,<,>",options="header " , ] | ewald summation and physically equivalent methods such as particle - mesh ewald , kubic - harmonic expansions , or lekner sums are commonly used to calculate long - range electrostatic interactions in computer simulations of polar and charged substances .
the calculation of pressures in such systems is investigated .
we find that the virial and thermodynamic pressures differ because of the explicit volume dependence of the effective , resummed ewald potential .
the thermodynamic pressure , obtained from the volume derivative of the helmholtz free energy , can be expressed easily for both ionic and rigid molecular systems . for a system of rigid molecules , the electrostatic energy and the forces at the atom positions
are required , both of which are readily available in molecular dynamics codes .
we then calculate the virial and thermodynamic pressures for the extended simple point charge ( spc / e ) water model at standard conditions .
we find that the thermodynamic pressure exhibits considerably less system size dependence than the virial pressure . from an analysis of the cross correlation between the virial and thermodynamic pressure ,
we conclude that the thermodynamic pressure should be used to drive volume fluctuations in constant - pressure simulations . |
we thank d. saumon for helpful discussions , and making available unpublished calculations , and d.j .
stevenson for constructive referee comments . this work was supported by the european community through a training and mobility of researchers grant to t.g .
\1989 . abundances of the elements : meteoritic and solar . _ geochim .
cosmochim . acta _ * 53 * , 197 - 214 . 1995 . solar models with helium and heavy elements diffusion .
modern physics _ * 67 * , 781808 . 1992 .
the molecular - metallic transition of hydrogen and the structure of jupiter and saturn . _ _ * 391 * , 817826 .
the abundance and distribution of water vapor in the jovian troposphere as inferred from voyager iris observations .
_ _ * 388 * , 648 - 688 . 1984 . the helium abundance of saturn from voyager measurements . _ _ * 282 * , 807 - 815
the c / h ratio in jupiter from the voyager infrared investigation . _ _ * 257 * , 901912 .
are the giant planets fully convective ? _ icarus _ , * 112 * , 337353 .
non - adiabatic models of jupiter and saturn . _ icarus _ , * 112 * , 354367 .
condensation of methane , ammonia and water and the inhibition of convection in giant planets .
_ science _ * 269 * , 16971699 . 1982 .
effects of differential rotation on the gravitational figures of jupiter and saturn .
_ icarus _ * 52 * , 509515 . 1989 .
structure and composition of giant planets interiors . in _ origin and evolution of planetary and satellite atmospheres _ ,
s. k. atreya , j. b. pollack , and m. s. matthews , eds . , university of arizona press , tucson , pp .
539563 . 1989 .
optimized jupiter , saturn and uranus interior models .
_ icarus _ * 78 * , 102118 .
1989 . a new analysis of the jovian 5-@xmath36 m voyager iris spectra .
_ icarus _ * 77 * , 457468 .
molecular dissociation in hot , dense hydrogen .
_ * 76 * , 12401243 .
formation of the giant planets .
phys . _ * 64 * , 544557 .
seismological observations with a fourier transform spectrometer : detection of jovian oscillations . _ _ * 267 * , 604622 .
metallization and electrical - conductivity of hydrogen in jupiter .
_ science _ * 273 * , 936938 .
comets , impacts and atmospheres .
_ icarus _ * 116 * , 215226 .
planetesimal dissolution in the envelopes of the forming giant planets .
_ icarus _ * 67 * , 409443 .
prediction of neon depletion in jupiter s atmosphere _
eos _ * 76 * , 343 .
an equation of state for low - mass stars and giant planets .
_ astrophys .
ser . _ * 99 * , 713741 .
structure of jupiter : galileo probe measurements . _ science _ * 272 * , 844845 .
interiors of the giant planets .
earth planet .
sci _ * 10 * , 257295 .
cosmochemistry and structure of the giant planets and their satellites .
_ icarus _ * 62 * , 415 . 1990 .
aneos analytic equations of state for shock physics codes , sandia natl . lab .
sand89 - 2951 .
metallization of fluid molecular - hydrogen at 140 gpa ( 1.4 mbar ) .
lett . _ * 76 * , 1860 - 1863 . 1938 . on the state of matter in the interior of planets .
_ astrophys .
j _ * 87 * , 508516 .
1996 . the helium mass fraction in jupiter s atmosphere .
_ science _ * 272 * , 849851 .
interior structure of the earth and planets ( translated by w.b .
hubbard & r.a .
masteler ) .
harwood academic pub .
modern models of giant planets . in _ high pressure research : application to earth and planetary sciences _ ( y. syono and m.h .
manghgnani , eds . ) , pp . 393401 .
* figure 1 * : + density profiles in models of jupiter with a fixed composition ( y=0.30 ) , using an adiabatic temperature profile and the @xmath8-eos ( plain line ) , adiabatic and ppt - eos ( plain line with discontinuity ) , non - adiabatic @xmath8-eos ( dashed line ) , and non - adiabatic ppt - eos ( dotted line ) . the thin plain and dashed lines correspond to adiabatic and non - adiabatic 2.5 mbar ppt - eos , respectively ( see text ) .
the temperatures of the models range between 2050 and 2450k at a pressure of 10kbar , 53006300k at 1 mbar , 1000013000k at 10 mbar and 1400021000k at 30 mbar .
upper curves are @xmath37k density profiles for water ice and olivine ( from thompson 1990 ) .
the dashed region represents the assumed uncertainty on the eos for heavy elements ( @xmath38 ) . within this region
, the continuous line corresponds to our `` preferred '' profile for @xmath25 .
_ inset _ : differences of the decimal logarithm of the jupiter density profiles with the same profile using the @xmath8-eos and an adiabatic structure .
the presence of a radiative region leads to a colder and denser model in the kbar region and deeper .
the presence of a ppt yields a denser metallic region . :
+ constraints on jupiter s @xmath39 ( mass of the central core ) , @xmath40 ( total mass of heavy elements ) , @xmath41 ( mass fraction of heavy elements in the molecular envelope ) , and @xmath24 ( mass fraction of heavy elements in the metallic envelope ) from interior models .
each colored surface represents models calculated with a given hydrogen - helium eos : ppt - eos ( red ) , ppt - eos with ppt at 2.5 mbar ( green ) , @xmath8-eos ( blue ) ( see text ) .
the central regions with a slightly lighter hue have been calculated with the `` preferred '' heavy elements eos ( see figure 1 ) , and can be used to estimate the uncertainty related to the eos of heavy elements ( @xmath25 ) .
plain and hatched surfaces correspond to adiabatic and non - adiabatic temperature profiles , respectively .
arrows indicate the magnitude and direction of uncertainties on @xmath42 ( mass fraction of ices in the core , from 0.5 to 1.0 ) , @xmath28 ( mass fraction of helium in the protosolar nebulae , from 0.275 to 0.28 ) and @xmath43 .
the latter corresponds to an increase by 1@xmath7 of @xmath44 , and to the assumption that jupiter rotates on cylinders and not as a solid planet .
the second effect is responsible for about half of the uncertainty , but only in one direction . for clarity ,
arrows are only shown for the ppt - non adiabatic models .
the mass mixing ratios of c , n , o and all other elements ( except h and he ) in abundances such that c / h=3@xmath45(c / h)@xmath46 , n / h=3@xmath45(n / h)@xmath46 , o / h=3@xmath45(o / h)@xmath46 , _ ... etc _ ( the solar abundances are from anders & grevesse 1989 ) are shown in the bottom right corner of the figure . the value of @xmath17 for any other composition can be inferred from this `` ruler '' . the molecular envelope represents about 15% ( 20% in the case of the 2.5 mbar ppt eos ) of the total mass of the planet ( @xmath47 ) .
we used @xmath35=0.0192 . | * using the helium abundance measured by galileo in the atmosphere of jupiter and interior models reproducing the observed external gravitational field , we derive new constraints on the composition and structure of the planet . we conclude that , except for helium which must be more abundant in the metallic interior than in the molecular envelope , jupiter could be homogeneous ( no core ) or could have a central dense core up to 12@xmath0 .
the mass fraction of heavy elements is less than 7.5 times the solar value in the metallic envelope and between 1 and 7.2 times solar in the molecular envelope . the total amount of elements other than hydrogen and helium in the planet is between 11 and 45@xmath0 .
* : _ icarus _ number of manuscript pages : 15 + number of tables : 0 + number of figures : 2 + keywords : jupiter , + interiors , planets + chemical abundances running head : composition of the interior of jupiter as shown by wildt ( 1938 ) , jupiter is mainly composed with hydrogen and helium .
however , the question of the abundance and partitioning of the other species ( `` heavy elements '' ) throughout the planet is still unsolved .
answering this question would yield much better constraints on models of formation of the giant planets .
it would also give insight into the behavior of chemical species at high pressures and generally into physical processes in the planet s interior .
a precise determination of the seismic properties of the planet would certainly be the best way to determine accurately its internal composition .
global oscillations of jupiter seem to have been detected ( mosser et al .
1993 ) , but the accuracy of the measurements have yet to be improved and the vibration modes more clearly identified . to date , constraints on the interior structure of the planet can be provided solely by interior models matching the observed gravitational field ( i.e. the radius , mass and gravitational moments @xmath1 , @xmath2 and @xmath3 ) and surface conditions ( temperature , luminosity , atmospheric helium abundance ) .
studies of the interior of giant planets ( hubbard & marley 1989 ; zharkov & gudkova 1992 ; chabrier et al .
1992 ; guillot et al .
1994b ) have generally been focused on calculating one possible interior model rather than calculating the ensemble of all possible models .
it is therefore difficult , from these models , to give constraints on the amount or distribution of elements in the interior of the planet .
furthermore , the new determination of the abundance of helium in the upper troposphere by the galileo probe ( niemann et al .
1996 ) requires a revision of these quantities .
previous calculations using the voyager data indicated that : @xmath4 ( gautier et al .
1982 ; conrath et al .
1984 ) , where x and y are the mass fractions of hydrogen and helium , respectively .
the mass fraction _ directly _ measured by the probe is now : @xmath5 ( von zahn & hunten 1996 ) .
these values have to be compared to the protosolar value estimated to @xmath6 ( bahcall and pinsonneault , 1995 ) . since no significant escape of the original
hydrogen and helium could occur over the lifetime of the planet , the interior of the planet must then be richer in helium than the upper regions in order to insure that the global hydrogen / helium ratio is equal to the protosolar one
. this conclusion holds for both the voyager and galileo measurements . however , the higher atmospheric helium mass fraction measured by galileo results in a smaller abundance of heavy elements in this region than previously estimated .
we have calculated new optimized jupiter models using the method described by guillot et al .
( 1994b ) .
we refer the reader to that article for details on the calculations and references for the data used .
we have used the galileo helium / hydrogen ratio , and have included the following uncertainties : 1 . on the equations of state ( eos ) , for hydrogen , helium and heavy elements 2 .
on the temperature profile ( convective / radiative ) 3 . on the internal rotation ( solid / differential ) 4 . on the distribution of helium in the interior
.
these uncertainties are discussed in detail below .
in addition , we have included the observational 1-@xmath7 uncertainties on the gravitational moments , atmospheric helium abundance and protosolar helium abundance .
the equation of state of hydrogen and helium is certainly the most important source of uncertainty in the calculations .
a large part of the jovian interior encompasses a pressure / temperature region where interactions between particles are important , and molecules , atoms , ions and electrons coexist .
in such a region , the behavior of even the simplest molecule , hydrogen , is poorly understood .
it undergoes a phase _ transition _ from a molecular to a metallic phase at pressures of the order of one to a few megabar , but it is not known whether this transition is of first order ( i.e. includes a discontinuity of the specific entropy ) or not .
the characterization of the two phases is also not obvious : experimental and theoretical evidence suggest that , at pressures of two megabars or less , a fraction of the electrons become itinerant while molecules still exist , before hydrogen eventually reaches a monoatomic truly metallic state at higher pressures ( of the order of a few megabars ) .
uncertainties in the equation of state are thus both quantitative and qualitative . in this work ,
we choose to estimate these uncertainties by using two hydrogen - helium eoss calculated by saumon , chabrier & van horn ( 1995 ) .
the first one ( @xmath8-eos ) assumes that the transition is continuous and is calculated by smoothly interpolating between the low - pressure molecular fluid and the high - pressure ionized plasma .
their second eos ( ppt - eos ) is consistently calculated and predicts the presence of a first order plasma phase transition between molecular and metallic hydrogen , the so - called ppt .
this transition occurs around 1.7 mbar for a typical jovian interior profile . in the molecular region
, the densities calculated with these eoss along isentropes characteristic of the jovian interior agree with those experimentally determined by weir et al .
( 1996 ) to better than 2% ( saumon & xie , in preparation ) , even though the adiabatic temperature gradients ( and hence temperatures ) differ .
however , the electrical conductivities measured indicate a continuous transition around 1.4 mbar ( nellis et al . 1996 ) , and no ppt , at least for pressures lower than 1.8 mbar . on the other hand , monte - carlo simulations by magro et al .
( 1996 ) indicate that a ppt occurs at slightly larger pressures .
we have therefore considered a third case , using the ppt - eos , but displacing the ppt to @xmath9 mbar .
although the resulting eos is thermodynamically inconsistent in the 1.72.5 mbar region , this approach yields a well - behaved density profile ( see figure 1 ) and therefore provides another useful estimation of the uncertainty of the hydrogen - helium eos .
finally , we choose to ignore the complexity of the molecular / metallic hydrogen transition itself , but this is mostly a semantic issue , assuming that the corresponding uncertainty on the density profile is properly represented by the different eoss . * insert figure 1 here * the partitioning of elements in jupiter s atmosphere , and in particular of helium , represents the second main uncertainty .
there are two mechanisms affecting equilibrium partitioning of elements in jupiter s hydrogen .
stevenson ( 1982 ) showed that in the metallic - hydrogen phase , pressure - ionized elements with @xmath10 have limited solubility . however , only helium is abundant enough to be directly affected by such limited solubility ( noble gases such as ne might be indirectly affected by he - immiscibility roulston & stevenson , 1995 ) . the standard explanation of jupiter s ( slightly ) reduced atmospheric helium abundance relies on such a phase separation via immiscibility in the metallic phase at pressures just above the molecular - metallic transition .
on the other hand , if a true ppt is present as a first - order phase transition , then the gibbs phase rule requires that the concentrations of _ all _ elements in hydrogen be discontinuous across this boundary . such partitioning between the metallic and molecular phases of hydrogen
would be in _ addition _ to any possible phase separation of helium due to limited miscibility in the metallic phase , and would in principle affect all elements rather than helium alone .
however , helium and most heavy elements are expected to be partitioned preferentially into the molecular phase .
the galileo probe results for elemental concentrations do not particularly argue for the predominance of one partitioning mechanism over the other , but they do confirm that helium is not enriched in the molecular region .
we emphasize that a composition change due to the ppt or that caused by limited helium miscibility are expected to occur in the same pressure region and therefore can not be distinguished by our models .
were it not the case ( for example if a helium - rich core was formed at pressures much larger than a few megabars ) , a slightly wider range of ice / rock core masses ( see below ) would be found .
although it bears little consequences for the structure of the interior models themselves , our poor knowledge of the equation of state of heavy elements in jupiter directly affects the abundance of heavy elements that is inferred from these models in order to conserve the protosolar helium / hydrogen ratio . as described by chabrier et al .
( 1992 ) and guillot et al .
( 1994b ) , the mass fraction of heavy elements is obtained after optimization of a model using : @xmath11 where @xmath12 and @xmath13 are the equivalent and real mass fraction of helium , respectively , and @xmath14 , @xmath15 and @xmath16 are the density of pure hydrogen , pure helium , and heavy elements respectively , along a pressure - temperature profile .
note that @xmath17 is uniform ( independent of @xmath18 and @xmath19 ) only in the limit @xmath20 , or if the ratios @xmath21 and @xmath22 are independent of @xmath18 and @xmath19 .
these variations of @xmath17 having no physical meaning , we used a mass average over the molecular and metallic regions , @xmath23 and @xmath24 , respectively . in order to estimate the uncertainty on @xmath25 , we calculated a low , a high and an intermediate @xmath16 profile . in the case of the low one , we assumed a small mean molecular weight at low pressures , and interpolated to high pressures where we used a zero temperature eos for pure h@xmath26o ice adding thermal effects using zharkov ( 1986 ) .
the high @xmath16 was calculated using a larger mean molecular weight at low pressures , including refractory elements ( rocks ) above 10kbar , assuming a 10 times solar rocks / ices mass mixing ratio and interpolating with zero temperature eoss for olivine ( mg@xmath26sio@xmath27 ) and h@xmath26o at high pressures .
an intermediate profile was calculated in the same way , but assuming solar ices / rocks ratio . at the molecular / metallic transition
, @xmath25 was allowed to jump from the low to the high profile , and vice - versa .
the corresponding profiles are shown in figure 1 as a hatched region , which represents the uncertainty on @xmath25 , both due to the unknown composition and to the poorly known eoss of these elements .
the possibility that the interior temperature gradient strongly departs from adiabaticity yields another important source of uncertainty in the model .
the temperature profile retrieved by the galileo probe is close to an adiabat at least down to the 20 bar level ( seiff et al .
1996 ) but condensation could yield brief subadiabatic or superadiabatic events ( guillot 1995 ) .
the possibility that a radiative region exists in the kbar region ( guillot et al .
1994b ) represents the largest source of uncertainty by far .
this was estimated by calculating rosseland opacities as in guillot et al .
( 1994a ) , but using new available opacity data for h@xmath26o and h@xmath26s ( see marley et al .
1996 ) .
in other regions ( except at the ppt ) we assumed the temperature structure to be adiabatic , as any departure from adiabaticity in the interior ( e.g. due to a compositional gradient ) would be indistinguishable from uncertainties in the eos itself .
a comparison of the density profiles calculated from the different equations of state and temperature profiles is given in figure 1 .
the presence of a radiative region induces an uncertainty of the order of 5% on the density in the 10-kbar region of the planet .
the uncertainty in the h - he eos is small at both low pressures and at very high pressures , but reaches as much as 10% in the intermediate pressure range of around one to five or so megabars , a pressure range that involves around a third of the planetary mass .
the @xmath8-eos is , on average , significantly less dense than the ppt - eos , and therefore requires a larger amount of heavy elements in order to reproduce the observed mean density of the planet .
the presence of a radiative region has the opposite effect , as it leads to a cooler planet . as in guillot
et al .
( 1994b ) , we assumed that a dense central core was present and included possible variations in composition ( pure ices / pure rocks ) .
its mass was allowed to vary in order to fit the observational constraints .
the presence of such a core is important for constraining formation theories ( e.g. mizuno 1980 ) .
although it is often thought that its mass is constant after the hydrogen - helium envelope has been captured , it is quite possible that convection redistributes it towards external regions , therefore yielding a smaller core , if any ( see stevenson 1985 ) .
we emphasize that the presence of a well defined core is _ assumed _ , and is not a consequence of the observed gravitational field of the planet .
very similar results would be obtained by distributing the core over the inner half of the planet .
there is therefore little difference between heavy elements in the core and in the metallic region .
finally , a small uncertainty arises from the unknown rotational state of jupiter s interior . as shown by hubbard ( 1982 ) , assuming that the planet is differentially rotating on cylinders ( with angular velocities constrained by observations ) yields slightly different gravitational moments @xmath1 , @xmath2 and @xmath3 than those inferred from a solid rotation model .
this uncertainty was also included in the present work .
the observed gravitational moments provide no constraints on density changes which encompass small masses , in particular on the structure of regions of pressures lower than kbars .
similarly , no real constraints exist on the structure of the core or its composition ( hubbard 1989 ; guillot et al
. 1994b ) . moreover , we find that the resulting models are very _ insensitive _ to the location of the helium abundance discontinuity , as long as it is between about 0.5 and 5 mbar . on the contrary ,
the models are fairly sensitive to changes in the location of the ppt because it induces significant changes of the density profile ( see figure 1 ) .
the resulting constraints on the abundance and distribution of heavy elements in jupiter are given in figure 2 .
the colored surfaces represent possible solutions for given hydrogen - helium eoss and temperature profile . within each patch all other uncertainties were included , as discussed previously .
the most significant are : ( i ) the eos for heavy elements ; ( ii ) the value of @xmath2 ; ( iii ) the value of @xmath28 ; and ( iv ) the unknown composition of the core .
the observational error on the atmospheric helium abundance measured by the galileo probe is small , but it is interesting to note that calculations using the voyager value would have yielded values of @xmath23 larger by 2 or 3 solar values .
the difference between adiabatic and non - adiabatic models is smaller than in guillot et al .
( 1994b ) because improved h@xmath26o opacities have increased the rosseland opacities by a factor @xmath29 ( see the discussion of the problem of hot bands in guillot et al .
1994a ) . * insert figure 2 here * a first important result is that the mass of the core can be equal to zero . in that case , we find it even possible for jupiter to be homogeneous ( @xmath30 ) , if we except small - scale phenomena and variations of the abundance of helium .
this requires that the density of the metallic phase is not too large compared to that of the molecular phase ( @xmath8-eos ) , and an important enrichment in heavy elements ( 4 to 7 times the solar value ) . in the case of the more realistic ppt - eos
, we find that the presence of a central core is necessary .
for all cases , the core is relatively small : less than 12@xmath0 ( and less than 8.5@xmath0 in the case of a pure rock core ) .
again , we stress that the elements that form the core could be distributed over the inner half of the planet , and that the distinction between heavy elements in the core and heavy elements in the metallic region is not obvious . other constraints on the amount of heavy elements present in the planet
suffer mainly from our relatively poor knowledge of the equation of state of hydrogen ( and helium ) at pressures of around one to a few megabars .
figure 2 shows that the total mass of heavy elements ( core + metallic + molecular envelope ) lies between 11 and 45@xmath0 .
the upper limit corresponds to an enrichment of a factor 7.4 compared to the solar value , which is quite large .
we estimate that it is probable that the density in the metallic region lies closer to that predicted by the ppt - eos than by the @xmath8-eos , as in fact favored by saumon , chabrier and van horn ( 1995 ) .
it is thus more likely that @xmath31 .
similarly , we obtain that in the metallic region , @xmath32 , or , when we disregard the @xmath8-eos , @xmath33 .
the constraint found on the mass fraction of heavy elements in the molecular envelope , @xmath34 , can be compared to the measurements of galileo s mass spectrometer ( niemann et al . , 1996 ) .
those confirmed that jupiter s atmosphere is enriched in ch@xmath27 by a factor 2.9 compared to the solar value of anders and grevesse ( 1989 ) , and in h@xmath26s by about the same factor .
the n / h ratio has not yet been evaluated from galileo , but voyager infrared data suggest an enhancement of 2.2 to 2.4 ( lellouch et al . , 1989 ; carlson et al . , 1992 ) .
the case of h@xmath26o is not clear because it condenses relatively deep in the atmosphere , but it is probably at least in solar abundance .
fig .
2 shows that an enrichment in all heavy elements by a factor 2 or 3 in the molecular region is certainly consistent with all possible interior models .
reciprocally , it is possible to show from our models that , even though o is responsible for almost 50% of the value of @xmath35 , large values of the o / h ratio can be consistent with the observed gravitational moments .
for example , assuming that all other elements are enriched by a factor 3 in the molecular region , we infer that the enrichment in h@xmath26o can be as high as 12 times the solar value .
however , this requires a relatively hard ( low density ) h - he eos in the molecular region . using the ppt - eos only
, we find that it is more likely that h@xmath26o is less abundant than 8 times solar in the molecular envelope . in any case
, these results do not permit to choose between very different assumptions concerning the enrichment of jupiter s outer envelope by meteorites and comets ( e.g. pollack et al .
1986 ; owen & bar - nun 1995 ) .
it is difficult to infer a lower limit to the abundance of water in the molecular envelope of the planet , but it appears that it would be difficult to reconcile an abundance less than solar with the values of @xmath23 found , as this would require significant enrichments in all other elements .
most of the uncertainties in the results shown in figure 2 are due to the relatively poor knowledge of the equation of state of hydrogen in the megabar region .
recent improvements in shock - wave experiments ( nellis et al . 1996 ) let us hope that a more accurate representation of the molecular / metallic transition of hydrogen will be achieved in the near future .
these improvements could easily yield abundances of heavy elements in jupiter more accurate by 50% .
similarly , a more precise value of @xmath2 would lead to better constraints on the structure of the planet .
the optimal precision that should be attained on @xmath2 is about 3 times the actual one , as one is limited by the uncertain internal rotation rate of the planet ( see hubbard 1989 for a discussion ) . in order to be a useful constraint ,
@xmath3 should be measured with a 10 times higher accuracy .
even so , the problem of differential rotation is even more acute than for @xmath2 , and a detailed study of the internal rotation of the planet should be done in order for such a measurement to be useful .
finally , more accurate high pressure opacities and generally a better understanding of transport processes in the planet are also required in order to determine precisely the abundance and partitioning of elements in jupiter s interior . |
this presentation is a cartoon of the effective field theory ( eft ) of two and three nucleon systems as it emerged in the last three years , using a lot of words and figures , and a few cheats . for details ,
i refer to the bibliographies in @xcite as well as a recent review @xcite , and papers with j .- w .
chen , r.p .
springer and m.j .
savage @xcite , p.f .
bedaque @xcite , and f. gabbiani @xcite .
i will mainly concentrate on the theory in which pions are integrated out as explicit degrees of freedom , but also comment on the extensions to include pions .
b. holstein s talk @xcite provides an introduction to eft , and e. epelbaum s contribution @xcite investigates one of the two proposals to include pions in more detail .
effective field theory methods are largely used in many branches of physics where a separation of scales exists @xcite . in low energy nuclear systems ,
the scales are , on one side , the low scales of the typical momentum of the process considered and the pion mass , and on the other side the higher scales associated with chiral symmetry and confinement .
this separation of scales produces a low energy expansion , resulting in a description of strongly interacting particles which is systematic and rigorous .
it is also model independent ( meaning , independent of assumptions about the non - perturbative qcd dynamics ) : given that qcd is the theory of strong interactions and that chiral symmetry is broken via the goldstone mechanism , wilson s renormalisation group arguments show that there is only one local low energy field theory which originates from it : chiral perturbation theory and its extension to the many - nucleon system discussed here .
three main ingredients enter the construction of an eft : the lagrangean , the power counting and a regularisation scheme .
first , the relevant degrees of freedom have to be identified . in his original suggestion how to extend eft methods to systems containing two or more nucleons
, weinberg @xcite noticed that below the @xmath1 production scale , only nucleons and pions need to be retained as the infrared relevant degrees of freedom of low energy qcd .
because at these scales the momenta of the nucleons are small compared to their rest mass , the theory becomes non - relativistic at leading order in the velocity expansion , with relativistic corrections systematically included at higher orders .
the most general chirally invariant lagrangean consists hence of contact interactions between non - relativistic nucleons , and between nucleons and pions , with the first terms of the form @xmath2\;+
\;g_a n^{\dagger } \vec{a}\cdot\sigma n\;- \\ & & -\;c_0 ( n^{{\mathrm{t } } } p^i n)^{\dagger } \ ( n^{{\mathrm{t } } } p^i n)\ ; + \;\frac{c_2}{8 } \left[(n^{{\mathrm{t } } } p^i n)^{\dagger } ( n^{{\mathrm{t } } } p^i ( \stackrel{\scriptscriptstyle\rightarrow}{{\partial}}- \stackrel{\scriptscriptstyle\leftarrow}{{\partial}})^2 n)+ \mathrm{h.c.}\right ] + \dots,\nonumber\end{aligned}\ ] ] where @xmath3 is the nucleon doublet of two - component spinors and @xmath4 is the projector onto the iso - scalar - vector channel , @xmath5 .
the iso - vector - scalar part of the @xmath6 lagrangean introduces more constants @xmath7 and interactions and has not been displayed for convenience .
the field @xmath8 describes the pion .
@xmath9 is the chirally covariant derivative , and @xmath10 ( @xmath11 ) the axial ( vector ) pionic current .
the interactions involving pions are severely restricted by chiral invariance . as such
, the theory is an extension of chiral perturbation theory and heavy baryon chiral perturbation theory to the many nucleon system . like in its cousins , the coefficients of the low energy lagrangean
encode all short distance physics
branes and strings , quarks and gluons , resonances like the @xmath1 or @xmath12 as strengths of the point - like interactions between particles .
as it is not possible yet to derive these constants by solving qcd e.g. on the lattice , the most practical way to determine them is by fitting to experiment .
because the lagrangean ( [ ksw ] ) consists of infinitely many terms only restricted by symmetry , an eft may at first sight suffer from lack of predictive power .
indeed , as the second part of an eft formulation , predictive power is ensured by establishing a power counting scheme , i.e. a way to determine at which order in a momentum expansion different contributions will appear , and keeping only and all the terms up to a given order .
the dimensionless , small parameter on which the expansion is based is the typical momentum @xmath13 of the process in units of the scale @xmath14 at which the theory is expected to break down .
values for @xmath14 and @xmath13 have to be determined from comparison to experiments and are a priori unknown .
assuming that all contributions are of natural size , i.e. ordered by powers of @xmath13 , the systematic power counting ensures that the sum of all terms left out when calculating to a certain order in @xmath13 is smaller than the last order retained , allowing for an error estimate of the final result .
for extremely small momenta , the pion does not enter as explicit degree of freedom , and all its effects are absorbed into the coefficients @xmath7 . this effective nuclear theory with pions integrated out ( ent(@xmath15 ) )
was recently pushed to very high orders in the two - nucleon sector @xcite where accuracies of the order of @xmath16 were obtained .
it can be viewed as a systematisation of effective range theory with the inclusion of relativistic and short distance effects traditionally left out in that approach .
because of non - analytic contributions from the pion cut , the breakdown scale of this theory must be of the order @xmath17 . for the ent with explicit pions , we would suspect the breakdown scale to be of the order of @xmath18 , since the @xmath1 is not an explicit degree of freedom in ( [ ksw ] ) .
even if calculations of nuclear properties were possible starting from the underlying qcd lagrangean , eft simplifies the problem considerably by factorising it into a long distance part which contains the infrared - relevant physics and is dealt with by eft methods and a short distance part , subsumed into the coefficients of the lagrangean .
qcd therefore `` only '' has to provide these constants , avoiding full - scale calculations of e.g. bound state properties of two nuclear systems using quarks and gluons .
eft provides an answer of finite accuracy because higher order corrections are systematically calculable and suppressed in powers of @xmath13 .
hence , the power counting allows for an error estimate of the final result , with the natural size of all neglected terms known to be of higher order .
relativistic effects , chiral dynamics and external currents are included systematically , and extensions to include e.g. parity violating effects are straightforward . gauged
interactions and exchange currents are unambiguous .
results obtained with eft are easily dissected for the relative importance of the various terms . because only @xmath19-matrix elements between on - shell states are observables , ambiguities nesting in `` off - shell effects ''
are absent . on the other hand , because only symmetry considerations enter the construction of the lagrangean , efts are less restrictive as no assumption about the underlying qcd dynamics is incorporated .
hence the proverbial quib that `` eft parameterises our ignorance '' . in systems involving two or more nucleons , establishing a power counting is complicated because unnaturally large scales have to be accommodated : given that the typical low energy scale in the problem should be the mass of the pion as the lightest particle emerging from qcd , fine tuning is required to produce the large scattering lengths in the @xmath20 wave channels ( @xmath21 ) . since there is a bound state in the @xmath22 channel with a binding energy @xmath23 and hence a typical binding momentum @xmath24 well below the scale @xmath14 at which the theory should break down , it is also clear that at least some processes have to be treated non - perturbatively in order to accommodate the deuteron , i.e. an infinite number of diagrams has to be summed . for simplicity ,
let us first turn to the eft in which the pion is integrated out into the coefficients @xmath25 of the lagrangean . here , a way to incorporate this fine tuning into the power counting
was suggested by kaplan , savage and wise @xcite , and by van kolck @xcite . at very low momenta ,
contact interactions with several derivatives
like @xmath26 should become unimportant , and we are left only with the contact interactions proportional to @xmath27 .
the leading order contribution to nucleons scattering in an @xmath20 wave comes hence from four nucleon contact interactions and is summed geometrically as in fig .
[ fig : deuteronprop ] in order to produce the shallow real bound state , i.e. the deuteron .
@xmath28 how to justify this ?
dimensional analysis allows any diagram to be estimated by scaling momenta by a factor of @xmath13 and non - relativistic kinetic energies by a factor of @xmath29 .
the remaining integral includes no dimensions and is taken to be of the order @xmath30 and of natural size .
this scaling implies the rule that nucleon propagators contribute one power of @xmath31 and each loop a power of @xmath32 . assuming that @xmath33 the diagrams contributing at leading order to the deuteron propagator are indeed an infinite number as shown in fig .
[ fig : deuteronprop ] , each one of the order @xmath34 .
the deuteron propagator @xmath35 has the correct pole position and cut structure when one chooses @xmath36 .
@xmath27 becomes dependent on an arbitrary scale @xmath37 because of the regulator dependent , linear uv divergence in each of the bubble diagrams .
indeed , when choosing @xmath38 , the leading order contact interaction scales as demanded in ( [ scalingksw ] ) . as expected for a physical observable , the @xmath6 scattering amplitude becomes independent of @xmath37 , the renormalisation scale or cut - off chosen .
all other coefficients @xmath25 can be shown to be higher order , so that the scheme is self - consistent .
observables are independent of the cut - off chosen
. the linear divergence does not show in dimensional regularisation as a pole in @xmath39 dimensions , but it does appear as a pole in @xmath40 dimensions which we subtract following the power divergence subtraction scheme @xcite .
dimensional regularisation is chosen to explicitly preserve the systematic power counting as well as all symmetries ( esp .
chiral invariance ) at each order in every step of the calculation . at leading ( lo ) , next - to - leading order ( nlo ) and often even n3lo in the two nucleon system
, it also allows for simple , closed answers whose analytic structure is readily asserted .
power divergence subtraction moves hence a somewhat arbitrary amount of the short distance contributions from loops to counterterms and makes precise cancellations manifest which arise from fine tuning @xcite .
the power counting of the zero and one nucleon sector of the theory are fixed by chiral perturbation theory and its extension to the one baryon sector .
therefore , the question to be posed is : how does the power counting of the contact terms @xmath25 change above the pion cut , i.e. when the pion must be included as explicit degree of freedom ?
we do not know yet , but the following two promising proposals are on the market .
if pulling out pion effects does not affect the running of @xmath41 too much , one surprising result arises : because chiral symmetry implies a derivative coupling of the pion to the nucleon at leading order , the instantaneous one pion exchange scales as @xmath30 and is _ smaller _ than the contact piece @xmath42 .
pion exchange and higher derivative contact terms appear hence only as perturbations at higher orders .
the lo contribution in this scheme is still given by the geometric series in fig .
[ fig : deuteronprop ] . in contradistinction to iterative potential model approaches , each higher order contribution is inserted only once . in this scheme , the only non - perturbative physics responsible for nuclear binding is extremely simple , and the more complicated pion contributions are at each order given by a finite number of diagrams .
for example , the nlo contributions to the deuteron are the one instantaneous pion exchange and the four nucleon interaction with two derivatives , fig . [
fig : nloksw ] .
the constants are determined e.g. by demanding the correct deuteron pole position and residue .
this approach is known as the `` ksw '' counting of ent , ent(ksw ) @xcite .
+ if on the other hand , the power counting for @xmath27 is dramatically modified when one includes pions , both the @xmath41 interactions and the one pion exchange might have to be iterated . in his paper , weinberg @xcite therefore suggested to power count not the amplitude as is usually done in eft but the potential , and then to solve a schrdinger equation with a chiral potential , as pictorially represented in fig .
[ fig : weinberg ] .
e. epelbaum s talk @xcite gives more details on the results obtained so far in this approach .
@xmath43 both power countings are presently under investigation for consistency and convergence , and each has its advantages and shortcomings @xcite .
if both approaches are self - consistent , nature will tell us which it chose .
although in general process dependent , the expansion parameter is found to be of the order of @xmath44 in most applications , so that nlo calculations can be expected to be accurate to about @xmath45 , and n2lo calculations to about @xmath46 . in all cases ,
experimental agreement is within the estimated theoretical uncertainties , and in some cases , previously unknown counterterms could be determined .
the first example we turn to demonstrates not only how simple it is to compute processes involving both external gauge and exchange currents in a self - consistent way , but especially how powerful the power counting is in estimates of theoretical uncertainties .
using the power counting ent(ksw ) in which pions are perturbative , the elastic deuteron compton scattering cross section @xcite to nlo is parameter - free with an accuracy of @xmath45
. contributions at nlo include the pion graphs that dominate the electric polarisability of the nucleon @xmath47 from their @xmath48 behaviour in the chiral limit .
because they are nlo , the power counting predicts their contribution to be sizeable , namely @xmath49 to @xmath50 .
the comparison with experiment in fig . [ fig : compton ] shows good agreement and therefore confirms the hb@xmath51pt value for @xmath47 .
the deuteron scalar and tensor electric and magnetic polarisabilities are also easily extracted @xcite .
-deuteron compton scattering at incident photon energies of @xmath52 and @xmath53 in eft(ksw ) @xcite , no free parameters .
dashed : lo ; long dashed : nlo without the graphs that contribute to the nucleon polarisability ; solid curve : complete nlo result .
the estimate of the accuracy of the calculation at nlo ( @xmath54 ) is indicated by the shaded area.,title="fig:",scaledwidth=49.0% ] -deuteron compton scattering at incident photon energies of @xmath52 and @xmath53 in eft(ksw ) @xcite , no free parameters .
dashed : lo ; long dashed : nlo without the graphs that contribute to the nucleon polarisability ; solid curve : complete nlo result .
the estimate of the accuracy of the calculation at nlo ( @xmath54 ) is indicated by the shaded area.,title="fig:",scaledwidth=49.0% ] compton scattering has also been investigated in ent(weinberg ) @xcite and in ent(@xmath15 ) @xcite .
surprisingly , the results all agree within the theoretical uncertainty of the calculations even at relatively high momenta .
this seems to suggest that non - analytic pion contributions from meson exchange diagrams are small .
since all interactions permitted by the symmetries must be included into the lagrangean , eft dictates that in the three body sector , interactions like @xmath55 with unknown strength @xmath56 are present . turning again to ent(@xmath15 )
, we therefore have to ask at which order in the power counting they will start to contribute .
bedaque , hammer and van kolck @xcite found the surprising result that an unusual renormalisation makes the three body force of leading order in the triton channel . in order to understand this finding , let us first consider the lo diagrams which come from two nucleon interactions .
the absence of coulomb interactions in the @xmath0 system ensures that only properties of the strong interactions are probed .
all graphs involving only @xmath27 interactions are of the same order and form a double series which can not be written down in closed form . summing all `` bubble - chain '' sub - graphs into the deuteron propagator
, one can however obtain the solution numerically from the integral equation pictorially shown in fig .
[ fig : lofaddeev ] within seconds on a personal computer .
@xmath57 @xmath57 @xmath57 @xmath58 + @xmath59 @xmath57 because one nucleon is exchanged in the intermediate state , each diagram is of the order of the nucleon propagator , i.e. @xmath60 , while the first three buddy force ( [ threebodyterm ] ) seems to be of order @xmath30 .
we are therefore tempted to assume that three body forces are at worst n2lo , i.e. of the order @xmath45 .
in the quartet channel , power counting suggests even n4lo because the pauli principle forbids three body forces without derivatives . however , such nave counting was already fallacious in the two body sector because of the presence of a low lying bound state and of a linear divergence in each of the bubble graphs making up fig .
[ fig : deuteronprop ] .
we therefore investigate more carefully the uv behaviour of the amplitude @xmath61 at half off - shell momenta @xmath62 .
the integral equation simplifies then to @xmath63\;\;,\end{aligned}\ ] ] where @xmath64 in the spin quartet channel , and @xmath65 in the doublet .
@xmath66 is the legendre polynomial of the second kind , @xmath67 the angular momentum of the partial wave investigated .
the equation is easily solved by a mellin transformation , @xmath68 , and indeed in most channels @xmath69 is real so that only one solution exists which vanishes at infinite momentum and hence is cut - off independent .
however , the fact that the kernel of ( [ uvfaddeev ] ) is not compact makes one suspicious whether this is always the case .
indeed , there exists one and only one partial wave in which two linearly independent solutions are found : the doublet @xmath20 wave ( triton ) channel , where @xmath70 . therefore in this channel ,
any superposition with an arbitrary phase @xmath71 is also a solution , and we find that @xmath72}{p}\;\;.\ ] ] because each value of the phase provides a different boundary condition for the solution of the full integral equation fig .
[ fig : lofaddeev ] , the on - shell amplitude @xmath73 will also depend crucially on @xmath71 .
this means that the on - shell amplitude seems sensitive to off - shell physics , and what is more , to a phase which stems from arbitrarily high momenta . a numerical study of the full half off - shell amplitude confirms these findings , fig .
[ fig : phillipsline ] .
this can not be . because physics must be independent of the cut - off chosen , this sensitivity of the on - shell amplitude on uv properties of the solution to ( [ fig : lofaddeev ] )
must be remedied by adding a counter term . and
since the power counting in the two nucleon sector is fixed , one can show that a necessary and sufficient condition to render cut - off independent results is to promote the three body force ( [ threebodyterm ] ) to lo , @xmath74 , and absorb all phase dependence into it , see fig .
[ fig : lotriton ] . the variation of @xmath75 with the cut - off or phase is known analytically from ( [ uvsolution ] ) , but its initial value is unknown .
therefore , one physical scale @xmath76 must be determined experimentally .
this one , new free parameter explains why potential models which provide an accurate description of @xmath6 scattering can vary significantly in their predictions of the triton binding energy @xmath77 and three body scattering length @xmath78 in the triton channel , although all of them lie on a curve in the @xmath79 plane , known as the phillips line , fig .
[ fig : phillipsline ] .
determining @xmath80 by fixing the three body scattering length to its physical value , the triton binding energy is found in ent(@xmath15)to be @xmath81 at lo . in the triton channel
has a dramatic effect on the scattering length @xmath82 .
right : comparison of the ent prediction for the phillips line with results from various potential models .
figures from ref . @xcite.,title="fig : " ] in the triton channel
has a dramatic effect on the scattering length @xmath82 .
right : comparison of the ent prediction for the phillips line with results from various potential models .
figures from ref .
@xcite.,title="fig : " ] @xmath83 it must again be stressed that the three body force of strength @xmath75 was added not out of phenomenological needs .
it cures the arbitrariness in the off - shell and uv behaviour of the two body interactions which would otherwise contaminate the on - shell amplitude . just as the off - shell behaviour of the two body amplitude fig .
[ fig : lofaddeev ] , the strength @xmath75 is arbitrary , and only the sum of two and three body graphs is physically meaningful . summing the deuteron bubbles , each graph in the upper line of fig . [
fig : lofaddeev ] behaves like @xmath84 in the uv .
but the solution of the faddeev equation goes like @xmath85 with irrational ( or even complex ) @xmath86 @xcite and is hence more than the nave sum of graphs . the limit cycle thus encountered in the triton channel is a new renormalisation group phenomenon and also explains the efimov and thomas effects @xcite .
however , the phenomenon occurs only in the triton channel . in all other partial waves , three body forces
enter only at higher orders @xcite . in the three body sector , the equations to be solved in ent(@xmath15)and ent(ksw ) are computationally trivial and can furthermore be improved systematically by higher order correction which involve only ( partially analytic , partially numerical ) integrations , in contradistinction to many - dimensional integral equations arising in other approaches . a comparative study between the theory with explicit , perturbative pions ( ent(ksw ) ) and the one with pions integrated out
was performed @xcite in the spin quartet @xmath20 wave for momenta of up to @xmath87 in the centre - of - mass frame ( @xmath88 ) . as seen above ,
the two formulations are identical at lo .
because three body forces enter only at high orders , this channel is completely determined by two body properties at the first few orders and no new , free parameters enter .
the calculation with / without explicit pions to nlo / n2lo shows convergence : for example , the scattering length is @xmath89 , and at nlo with ( without ) perturbative pions @xmath90 ( @xmath91 ) . at n2lo ,
@xcite report @xmath92 , and the experimental value is @xmath93 .
comparing the nlo correction to the lo scattering length provides one with the familiar error estimate at nlo : @xmath94 .
the nlo calculations with and without pions lie within each other s error bar .
the n2lo calculation is inside the error ascertained to the nlo calculation and carries itself a theoretical uncertainty of about @xmath95 .
the calculation of pionic corrections in ent(ksw ) shows that they are although formally nlo indeed much weaker .
the difference to ent(@xmath15)should appear for momenta of the order of @xmath96 and higher because of non - analytic contributions of the pion cut , but those seem to be very moderate , see fig .
[ fig : delta ] .
this and the lack of data makes it difficult to assess whether the ksw power counting scheme to include pions as perturbative increases the range of validity over the pion - less theory .
@xcite and doublet @xmath97 @xcite wave phase shifts in @xmath0 scattering .
legend left : dashed : lo ; solid ( dot - dashed ) line : nlo with perturbative pions ( pions integrated out ) ; dotted : n2lo without pions .
realistic potential models : squares , crosses , triangles .
stars : @xmath98 phase shift analysis.,title="fig:",scaledwidth=52.0% ] @xcite and doublet @xmath97 @xcite wave phase shifts in @xmath0 scattering .
legend left : dashed : lo ; solid ( dot - dashed ) line : nlo with perturbative pions ( pions integrated out ) ; dotted : n2lo without pions .
realistic potential models : squares , crosses , triangles .
stars : @xmath98 phase shift analysis.,title="fig:",scaledwidth=44.0% ] finally , the real and imaginary parts of the higher partial waves @xmath99 in the spin quartet and doublet channel were found @xcite in a parameter - free calculation , see fig .
[ fig : delta ] . within the range of validity of this pion - less theory , convergence is good , and the results agree with potential model calculations ( as available ) within the theoretical uncertainty .
that makes one optimistic about carrying out higher order calculations of problematic spin observables like the nucleon - deuteron vector analysing power @xmath100 where eft will differ from potential model calculations due to the inclusion of three - body forces .
many questions remain open : does the power counting nature chose include pions perturbatively or non - perturbatively ? at the moment , some people advocate a mixture
. to investigate processes with pions in the initial or final state might be helpful @xcite . how to extend the analysis in the triton channel systematically to higher orders ?
technically , how to regularise a faddeev equation numerically with external currents coupled in a field theory ? how do four body forces scale ?
extend to nuclear matter !
u. van kolck , m. savage and r. seki , eds . , nuclear physics with effective field theory " , proceedings of the int - caltech workshop at caltech ( 1998 ) , world scientific ; p. bedaque , u. van kolck , m. savage and r. seki , eds . , nuclear physics with effective field theory ii " , proceedings of the int - caltech workshop at the int ( 1999 ) , world scientific . p.f .
bedaque , h.w . hammer and u. van kolck , phys .
82 * , 463 ( 1999 ) ; p.f .
bedaque , h.w . hammer and u. van kolck , nucl .
* a646 * , 444 ( 1999 ) ; p.f .
bedaque , h.w . hammer and u. van kolck , nucl .
* a676 * , 357 ( 2000 ) . | progress in the effective field theory of two and three nucleon systems is sketched , concentrating on the low energy version in which pions are integrated out as explicit degrees of freedom .
examples given are calculations of deuteron compton scattering , three body forces and the triton , and @xmath0 partial waves . fewfeyn |
vector mesons can be produced by photonuclear interactions in ultra - peripheral heavy ion collisions ( upc ) .
the electromagnetic interactions occur at impact parameters @xmath2 larger than twice the nuclear radius @xmath3 , where no hadronic interactions can occur@xcite . in these upcs , the electromagnetic field of one nucleus acts as almost - real photon field , following the weizscker - williams approach
. these photons can fluctuate into a quark - antiquark pair , which can scatter elastically from the other nucleus , emerging as a real vector meson .
the photons can also fluctuate into virtual @xmath4 pairs , with one of the pions scattering from the other nucleus and emerging as a real pion pair .
purely electromagnetic interactions also occur . at energy scales above @xmath5 , for most purposes ,
these may be described as two - photon interactions , although 3 ( or more ) photon interactions are also possible@xcite .
these reactions can produce @xmath1 , @xmath6 and @xmath7 pairs , scalar or tensor mesons , and meson pairs@xcite . since the photons couple to charge , the production rate is a sensitive test of the internal charge content of mesons , and can be used to rule out glueball candidates .
upcs occur at moderate impact parameters , @xmath8 where @xmath9 is the vector meson mass and @xmath10 the ( lab frame ) lorentz boost .
because the electromagnetic fields are so strong , additional photonuclear interactions may accompany the vector meson production .
in addition to the @xmath0 producing photon , the nuclei may exchange one or more additional photons which may excite the target nuclei into a giant dipole resonance or higher excitation . when the excited nuclei decay , they emit one or more neutrons .
we report the first observation of @xmath0 production in upcs , both with and without nuclear excitation .
we also observe @xmath1 pairs produced in upcs .
these reactions were observed in @xmath11 collisions at a center of mass energy @xmath12 gev at the relativistic heavy ion collider ( rhic ) by the solenoidal tracker at rhic ( star ) detector .
figure [ fig : event ] shows a typical event : two charged particles are visible in an otherwise empty detector .
the particles are roughly back - to - back , showing that the pair has a small transverse momentum @xmath13 .
[ fig : event ]
the rates for exclusive @xmath0 production may be calculated using data on @xmath14 collected from hera and fixed target experiments .
the photon flux is given by the the weizscker - williams virtual photon method .
the cross section to produce a vector meson @xmath15 , @xmath16 , is determined by a glauber calculation using the @xmath17 data as input .
the calculated @xmath16 cross sections@xcite agree with data from lower energy fixed target experiments .
for less - well studied mesons , the measured vector meson rates may be used to determine the meson - nucleon interaction cross sections . the production cross section at a given photon energy @xmath18 ( in the lab frame ) maps into the rapidity @xmath19 of the final state vector meson : @xmath20 .
a photon with a lab - frame energy @xmath21 is lorentz boosted to an energy @xmath22 in the target frame , so the @xmath23 center of mass energies are comparable to those reached at fermilab fixed - target experiments , somewhat below those reached at hera . the total cross section is the integral over @xmath18 of the photon flux times @xmath16 . for gold collisions at @xmath24
gev the @xmath0 production cross section is expected to be about 400 mb , 5% of the total hadronic cross section .
a @xmath25 final state may be produced directly or through the @xmath0 .
the amplitudes for @xmath0 production , @xmath26 , and direct @xmath25 production , @xmath27 , interfere and@xcite @xmath28 where the @xmath0 width is corrected for the increasing phase space as @xmath29 increases .
the @xmath0 component undergoes a 180@xmath30 phase shift at @xmath31 , so the interference skews the @xmath0 peak shape , enhancing production for @xmath32 and suppressing the spectrum for @xmath33 .
these upcs are characterized by final states with small @xmath13 because of the coherent coupling .
the meson @xmath13 comes from the photon , @xmath34 and the @xmath13 acquired in the scattering , about @xmath5 , added in quadrature .
vector meson production can occur at either of the nuclei ; because the strong force has a short range , the production must be inside or very near the nuclei . because it is impossible to determine which nucleus was the source , the amplitudes from the two production sites ( ions ) interfere , reducing the number of @xmath0 with @xmath35@xcite . because the short - lived @xmath0 decay before they travel the distance @xmath2 , the interference is sensitive to the post - decay wave function , and can be used for tests of quantum mechanics .
electron - positron pairs may be produced in upcs by purely electromagnetic processes .
although the coupling is large ( @xmath36 ) , in the kinematic range accessible to star , higher order effects are likely to be small , and the cross section is calculable based on the collision of two almost - real virtual photons@xcite .
photon exchange can also excite one or both nuclei@xcite .
collective excitations , such as the giant dipole resonance , are possible , along with higher energy incoherent photonuclear interactions involving hadron production .
calculations indicate that , in mutual coulomb dissociation , the two excitations are likely to occur independently , via two separate photons .
the cross section for mutual coulomb exchange is sizable , around 3.6 barns at @xmath37 gev@xcite .
this process may be used as a luminosity monitor .
mutual excitation also tags events with low impact parameters .
the impact parameter dependence depends only on the photon flux variation .
the flux at a given @xmath2 is given in the weizscker - williams formalism@xcite .
multiple photons from a single nucleus are emitted independently of each other@xcite .
as long as there is no interference between final states , the individual interactions are independent events , with the cross section for multiple interactions given by an integration of the joint probabilities over transverse space : @xmath38 .
\label{eq : factor}\ ] ] where @xmath39 is the probability of @xmath0 production and @xmath40 is the probability of mutual nuclear excitation .
there are some feynman diagrams for which factorization does nt apply .
for example , when photon emission excites the emitting nucleus , factorization fails .
these processes have been studied for two - photon upcs , and the non - factorizing amplitudes were found to be small@xcite .
it seems likely that a similar independence holds for photonuclear interactions .
the star detector reconstructs charged particles using a 4.2 meter long time projection chamber ( tpc)@xcite in a 0.25 t solenoidal magnetic field .
the tpc has an inner radius of 50 cm , and an outer radius of 2 m. the charged particle pseudorapidity ( @xmath19 ) acceptance depends on the position of the interaction . for the data discussed here , the interactions were radially within a few cm of the center of the tpc , but spread longitudinally ( in @xmath41 ) , with @xmath42 cm .
tracks originating near @xmath43 ( the center of the tpc ) were tracked for @xmath44 .
the momentum resolution was about @xmath45 .
tracks with @xmath46 mev / c were reconstructed with good efficiency .
particles were identified by their energy loss ( @xmath47 ) in the tpc . for low - multiplicity final states ,
the @xmath47 resolution was about 8% .
the tpc is surrounded by a cylindrical central trigger barrel ( ctb ) . for tracks originating near @xmath43
, it is sensitive to @xmath48 .
this barrel consists of 240 scintillator slats , each covering @xmath49 by @xmath50 .
two zero degree calorimeters ( zdcs ) at @xmath51 meters from the interaction point detect neutrons from nuclear breakup .
these calorimeters have @xmath52 acceptance for single neutrons from nuclear breakup@xcite .
the initial trigger decision uses lookup tables and field programmable gate arrays to initiate tpc readout about 1.5 @xmath53s after the collision .
tpc readout took 10 msec , allowing for up to 100 events / sec to be read out ( at 100% deadtime ) .
a higher level filter uses on - line track reconstruction by a small processor farm @xcite to remove unwanted events with vertices outside the beam interaction region .
two separate triggers were used to study @xmath0 production . the first ( topology ) trigger selected events with a small number of tracks detected by the ctb .
it divided the ctb into 4 azimuthal quadrants , and required a hit in the opposing north and south quadrants .
events with hits in the top or bottom quadrants were vetoed as probable cosmic rays .
we took 7 hours of data with this trigger , recording about 30,000 events on tape .
the majority of the triggers were due to cosmic rays , beam - gas events , and debris from upstream interactions .
[ fig : pctrig ] our analysis selected events with a vertex with exactly 2 tracks within 2 cm of the centerline of the tpc , and with @xmath54 m. the track @xmath47 were required to be consistent with pions .
electrons and pions can be separated for @xmath55 mev / c . at higher
momenta the two @xmath47 bands overlapped . finally , to reject the remaining cosmic rays , we require the events to have a 3-dimensional opening angle greater than 3 radians ( i.e. they must not be perfectly back - to - back ) . fig .
[ fig : pctrig](a ) shows the pair @xmath13 spectrum of the selected unlike - sign ( dots - net charge 0 ) and like - sign pairs ( histograms ) .
the unlike pairs are strongly peaked for @xmath56 mev / c .
this is consistent with production that is coherent with both nuclei ; events with @xmath57 mev / c are considered our signal .
the like - sign pairs have no such enhancement , and serve as a background sample .
the like - sign pairs have been normalized to match the unlike - sign in the signal - free region 250 mev @xmath58 500 mev ; this entailed scaling them up by a factor of 2.1 .
[ fig : pctrig](b ) shows the invariant mass of the pairs with @xmath57 mev / c .
the points are the unlike - sign events , while the hatched histogram are the scaled like - sign pairs .
the like - sign pairs are concentrated at relatively low masses , while the net charge 0 pairs are peaked around the @xmath0 mass .
the asymmetric peak is well fit by eq .
( 1 ) , with @xmath59 similar to that found by the zeus collaboration for @xmath60@xcite@xcite .
however , some of the events in the low - mass shoulder are likely due to @xmath1 pairs ; subtracting this background will decrease @xmath59 .
there is no evidence for any neutrons in the zdc data from these events .
data on @xmath0 production accompanied by mutual nuclear excitation were collected with the minimum bias trigger .
this trigger required that signals from one or more neutrons be detected in each zdc .
the event selection for this analysis was the same as for the topologically triggered sample .
[ fig : mbtrig]a shows the @xmath13 distribution of the track pairs , and fig .
[ fig : mbtrig]b shows the invariant masses of the @xmath25 combinations with @xmath62 mev / c .
the overall background is slightly lower than for the topological trigger , but the composition is quite different . here , the like - sign pair background was scaled up by a factor of 2.3 to match the like - sign pairs in the region 250 mev / c @xmath63 500 mev / c . compared to the topologically triggered sample , the backgrounds from beam gas events and cosmic rays are greatly reduced , but the contribution due to grazing nuclear collisions is larger . despite these differences , the @xmath0 mass peak has a similar shape and @xmath59 as the topologically triggered data .
[ fig : mbtrig ]
this analysis selected events where both particles were identified as leptons ( and hence with @xmath64 mev / c ) .
leptons from these are shown by the triangles in fig .
[ fig : electrons]a . fig .
[ fig : electrons]b shows the @xmath13 spectrum for events with two identified electrons with @xmath65 mev / c .
there is a clear peak for pair @xmath66 mev / c .
the photon @xmath67 ( considerably less than the @xmath68 sometimes quoted ) , so the @xmath13 spectrum is consistent with the expected photon @xmath13 .
[ fig : electrons ]
we have observed for the first time the reactions @xmath69 , @xmath70 and @xmath71 .
the final states have a small perpendicular momentum , showing their coherent coupling to both nuclei . in 2001
, rhic is running at @xmath72 gev and will attempt to reach full design luminosity , with shorter beam bunches ( @xmath73 cm ) .
star has added several new detector elements : a silicon vertex tracker and forward tpcs , along with an electromagnetic calorimeter .
in addition , the star trigger will be able to take topological triggers in parallel with central collision triggers .
together , these improvements should give us several orders of magnitude more data .
we expect to study additional final states , including the @xmath74 and the @xmath75 final states of the @xmath76 and @xmath77 , and to study @xmath0 production with greater precision and increased angular acceptance .
this will allow us to study the interference between the two vector meson production sites .
we will also begin to study two - photon production of scalar and tensor mesons . | we report the first observation of exclusive @xmath0 production in ultra - peripheral collisions at rhic .
the @xmath0 are produced electromagnetically at large impact parameters where no hadronic interactions occur .
the produced @xmath0 have a small perpendicular momentum , consistent with production that is coherent on both the photon emitting and scattering nuclei .
we observe both exclusive @xmath0 production , and @xmath0 production accompanied by electromagnetic dissociation of both nuclei .
we discuss models of vector meson production and the correlation with nuclear breakup .
we also observe @xmath1 pair production in these ultra - peripheral collisions .
address=70 - 319 lbnl , berkeley , ca , 94720 usa , [email protected] , thanks = this work was commissioned by the aip presented at inpc 2001 , july 30-aug . 3 , 2001 , berkeley , ca . |
the type ia supernovae ( sne ia ) are observed to be fainter than expected from the luminosity - redshift relationship in a decelerating universe .
this unanticipated dimming was first attributed to an accelerating expansion of the universe @xcite .
although the existence of cosmic acceleration has been verified by several other observations , initially , there had been some debates on the interpretation of underlying physical mechanism for the observed sne ia dimming .
for example , dust in the milk way and oscillation of photons propagating in extragalactic magnetic fields into very light axions had been proposed to account for the dimming @xcite .
any kind of photon number violation , such as absorption , scattering or axion - photon mixing , sensibly imprints its influence on the tolman test @xcite , which can be rewritten as a relationship among cosmological distance measurements known as the famous distance - duality ( dd ) relation @xcite , @xmath8 where @xmath9 is the redshift , @xmath0 and @xmath1 are the luminosity distance and the angular diameter distance ( add ) respectively .
this reciprocity relation holds for general metric theories of gravity in any background , not just in that of the friedmann - lem@xmath10itre - robertson - walker background , and it is also valid for all cosmological models based on the riemannian geometry .
that is , its validity depends neither on einstein field equations nor on the nature of the matter - energy content .
the dd relation plays an important role in modern cosmology @xcite , and , in most cases , it has been applied , without any doubt , to analyze the cosmological observations .
however , the reciprocity relation may be violated if photons do not travel on null geodesics or the universe is opaque .
fortunately , it is in principle possible to perform a valid check on the dd relation by means of astronomical observations .
the basic idea is to search for observational candidates with known intrinsic luminosities as well as intrinsic sizes , and then determine their @xmath0 and @xmath1 to test the etherington relation directly .
it is difficult for us to find objects of the same class with both known intrinsic luminosities and intrinsic sizes .
thus , a @xmath6cdm cosmological model is usually assumed when one performs tests by utilizing observed @xmath0 or @xmath1 @xcite and the results show that there is no strong evidence for deviations from the standard dd relation .
recently , a model - independent method has been proposed to test dd relation by considering two different classes of objects , for example , sne ia and galaxy clusters , from which @xmath0 and @xmath1 are determined separately @xcite . for a given add data , in order to obtain the corresponding @xmath0 from sne ia , a selection criteria @xmath11 is adopted . using the phenomenological parameterized forms @xmath12 with @xmath13 and @xmath14 , and the data from union2 sn ia @xcite and galaxy clusters , it was found that the dd relation can be accommodated at @xmath15 c. l. for the elliptical @xmath5 model @xcite and at @xmath16 c. l. for the spherical @xmath5 model @xcite .
more recently , in order to avoid the bias from the redshift difference between @xmath0 and add , methods , such as , the binning of sne ia @xcite , the interpolation @xcite and local regression @xcite from nearby sne ia points for a given galaxy cluster , were proposed and similar results were obtained . in addition , the dd relation tests by use of sne ia and gas mass function of galaxy clusters were carried out and similar results were also obtained @xcite .
so , overall , all the tests performed so far show that there is no strong indication of the dd relation violation . however , let us note that systematic uncertainties resulting from the morphological models of galaxy clusters and the redshift difference between @xmath0 and add might exert influences on dd relation tests .
if one considers that the photon traveling along null geodesic is unassailable , the dd relation violation most likely implies non - conservation of the photon number which has a mundane origin ( scattering from dust or free electron ) or an exotic origin ( photon decay or photon mixing with other light states such as the dark energy , dilaton or axion @xcite ) . in this case , the flux received by the observer will be reduced and so the universe is opaque .
if we assume that the flux from the source is decreased by a factor @xmath17 , then the inferred ( observed ) luminosity distance differs from the true " one @xcite @xmath18 where @xmath19 is the opacity parameter which denotes the optical depth associated to the cosmic absorption .
initially , more _
et al . _
@xcite studied the cosmic opacity by examining the difference of the opacity parameter at redshifts @xmath20 and @xmath21 , @xmath22 , where the difference of the observational luminosity distance ( @xmath23 ) at these two redshifts was estimated from two subsamples of essence sn ia @xcite and the corresponding @xmath24 was derived from the distance measurements of baryonic acoustic feature @xcite in the context of @xmath6cdm . assuming flat priors on @xmath25 and @xmath26 in the range of @xmath27 , and uniformly spaced values of @xmath28
$ ] , they found that a transparent universe is favored ( posterior probabilities of @xmath29 peaked at 0 ) and @xmath30 at 95% c. l .. this method has been applied to investigate the homogeneity of the cosmic opacity in different redshift regions and the results suggest that the cosmic opacity oscillates between zero and non - zero values as redshift varies @xcite .
later , avgoustidis _ et al . _
@xcite carried out further studies by assuming an optical depth parameterization @xmath2 or @xmath3 for small @xmath31 and @xmath32 .
there they took the standard luminosity distance in the spatially flat @xmath6cdm ( @xmath33 ) and the union sn ia @xcite for @xmath34 and @xmath35 respectively .
in addition to the sne ia data , they also used the measurements of the cosmic expansion @xmath36 @xcite . by taking @xmath37 $ ] ,
@xmath38 $ ] and @xmath39 $ ] @xcite all uniformly spaced over the relevant intervals in a flat @xmath6cdm model and performing a full bayesian likelihood analysis , they obtained a result @xmath40 ( @xmath41 c. l. ) , which corresponds to an opacity @xmath42 ( 95% c. l. ) for the redshift range between 0.2 and 0.35 , almost a factor of 11 stronger than the constraint obtained in ref .
recently , lima _ et al . _
@xcite reexamined this issue by confronting the luminosity distance which is dependent on two free parameters , i.e. , the so called cosmic absorption parameter ( @xmath43 ) and the matter density ( @xmath26 ) , with observations , using a subsample of union2 sn ia obtained by selecting sne ia with redshifts greater than @xmath44 in order to avoid effects from hubble bubble , and they found that the einstein - de sitter model ( @xmath45 ) could be allowed at 68.3% ( 95.4% ) c. l. in the case of a constant ( epoch - dependent ) absorption and concluded that a cosmic absorption may be responsible for the dimming of the distant sne ia without the need of an accelerated expansion of the universe .
however , all these studies concerning the cosmic opacity assume a ( flat ) @xmath6cdm model , and are thus model - dependent . here
we propose another model - independent method to examine the cosmic opacity and investigate its possible implications for the cosmic evolution .
if we assume that the violation of the dd relation is purely caused by the photon number non - conservation , then we can find out whether the universe is opaque by checking the possible violation of the dd relation .
it should be emphasized that the cosmic absorption not only affects the luminosity distance measurements of sne ia observations as shown in eq .
( [ eq3 ] ) , but also exerts influences on the angular diameter distance measurements determined from sze+x - ray surface brightness observations @xcite , @xmath46 where @xmath47 is the temperature change due to the sunyaev - zeldovich effect ( sze ) when the cosmic microwave background ( cmb ) photons pass through the hot intracluster medium and @xmath48 is the x - ray surface brightness of galaxy clusters .
the sze spectra distortion of the cmb is determined by measuring the intensity decrements , @xmath49 , which is sensitive to the cosmic absorption .
additionally , the surveys of x - ray surface brightness are also sensitive to the opacity of the universe
. supposing the absorption is frequency independent in the observed frequency range ( from microwave band to x - ray band ) , the true " add connects the observed one measured in an opaque universe with @xmath50 .
thus , in actual calculations , the dd relation takes the following form : @xmath51 now we will use the the largest union2.1 sn ia sample @xcite and the add data from galaxy cluster samples @xcite to test , model - independently , the possible violation of the dd relation , which can be translated to a possible cosmic opacity .
the observed @xmath0 and @xmath1 come from the latest union2.1 sn ia and galaxy clusters samples @xcite , respectively .
actually , there are a number of inherent uncertainties in the selected astrophysical objects from which the observed @xmath1 are derived , e. g. , the cluster asphericity @xcite and the model for the cluster gas distribution @xcite . in this paper , we consider the elliptical @xmath5 model galaxy clusters sample @xcite , spherical @xmath5 model , @xmath52-cut spherical @xmath5 model and hydrostatic equilibrium model galaxy clusters samples @xcite to investigate the impact of these inherent uncertainties on the cosmic opacity test .
in order to place constraints on the cosmic opacity parameter @xmath19 , we first parameterize it with two monotonically increasing functions of redshift , i.e. , @xmath2 and @xmath3 @xcite .
these two parameterizations are basically similar for @xmath53 but they differ when @xmath9 is not very small . as the data applied in our following analysis discretely distribute in the redshift range @xmath54 , our analysis we are going to perform
may tell us something about the possible dependence of the test results on the parametric forms for @xmath19 . to obtain @xmath55 determined by the following expression : @xmath56,\ ] ] the data pairs of observed @xmath0 and @xmath1 almost at the same redshift should be supplied . for the observed @xmath0 , the largest union2.1 sn ia is considered .
galaxy cluster samples , where the @xmath1 are obtained by combining the sze+x - ray surface brightness measurements @xcite , are responsible for providing the observed @xmath1 . the first one , including a selection of 7 clusters compiled by mason _ et al .
@xcite and a sample of 18 clusters collected by reese _
et al . _
@xcite , was re - analyzed by filippis _
_ @xcite by assuming an elliptical @xmath5 model for the galaxy clusters .
the second kind of samples are compiled by bonamente _
et al . _
@xcite with three different models for the cluster plasma and dark matter distribution , i.e. , the spherical @xmath5 model , spherical @xmath5 model with @xmath57 kpc cut , and hydrostatic equilibrium model .
therefore , the data derived from these three different models are adopted to check whether the cosmic opacity tests are sensitive to the model for the cluster gas distribution .
the observed @xmath0 are binned from the data points of union2.1 sn ia with their redshifts satisfying the certain criteria ( @xmath58 ) to match the observational data of the add samples @xcite .
this binning method can minimize the statistical errors originating from the redshift difference between @xmath0 and @xmath1 .
on the other hand , we alter @xmath59 from 0.000 to 0.005 to ensure the number of the clusters that share the same sne to be as few as possible , so as to reduce the dependence of opacity tests on the correlation of redshift - matched sne . after obtaining @xmath55 from these selected data pairs
, we estimate the free parameters of a given parametric form by using the standard minimum @xmath60 route : @xmath61 ^ 2}{\sigma^2_{\tau_{\mathrm{obs}}}},\ ] ] where @xmath62 is the error of @xmath63 associated with the observed @xmath0 and @xmath1 , and @xmath64 represents the free parameters to be constrained .
the graphic representations and numerical results of the probability distribution of the opacity parameter @xmath31 constrained from the model - independent tests are shown in figures ( [ fig1 ] , [ fig2 ] ) and table ( [ tab1 ] ) .
these suggest that the dependence of test results on the above - chosen parameterizations for @xmath65 is relatively weak .
similar to the results obtained by examining the cosmic opacity in a particular redshift range ( 0.20 - 0.35 ) @xcite and deforming the dd relation in terms of the cosmic absorption parameter @xcite , we find , from figure ( [ fig1 ] ) , that an almost transparent universe is also favored by the elliptical @xmath5 model galaxy clusters sample @xcite . for the add samples given by bonamente _
et al . _
@xcite , the results are shown in figure [ fig2 ] .
we find that the results are nearly not sensitive to the model of cluster gas distribution and a transparent universe can only be marginally accommodated at 95.4% c. l. ( even at 99.7% c. l. when the @xmath57 kpc - cut spherical @xmath5 model is considered ) .
that is , all the constraints on the opacity parameter obtained from the bonamente _
_ samples prefer an opaque universe .
these results are clearly different from what were obtained based on the @xmath6cdm in refs .
@xcite , where a transparent universe is obviously supported .
in fact , these cosmic opacity test results are very similar to the previous model - independent tests for the dd relation @xcite .
however , our objective here is the cosmic opacity test with an assumption that the violation of the dd relation entirely originates from the non - conservation of photon number , rather than the dd relation test itself . in order to explore the implications of the cosmic opacity ,
let us transform the sne ia distance modulus in a transparent universe into that in an opaque one @xmath66\tau(z)\;,\ ] ] and study the cosmological constraints resulting from this correction .
since the high redshift galaxy cluster data are absent in our discussion of the cosmic opacity , the distance - modulus - modified sne ia used to investigate the cosmological constraints are cut down from union2.1 with the criteria @xmath67 and @xmath68 when clusters in the filippis _
_ sample and the bonamente _
_ samples are applied , respectively .
since the dependence of cosmic opacity tests on the above - chosen parametric forms of @xmath65 is relatively weak , we only consider results obtained from the linear parametrization ( @xmath2 ) in the following cosmological implication analysis .
for the flat @xmath6cdm model , different from the methods used in refs .
@xcite , we examine the probability distributions of @xmath26 by considering the possible ( 68.3% c. l. range ) opacity parameter @xmath31 constrained from the previous model - independent tests .
the results are shown in figures ( [ fig3 ] , [ fig4 ] ) .
we find that the opacity parameter @xmath31 constrained from previous cosmological - model - independent tests impact slightly on the likelihood distributions of @xmath26 and a universe with @xmath69 is required to account for the dimming of the sne ia .
this differs from the results in ref .
@xcite , where the einstein - de sitter universe ( @xmath45 ) can be easily accommodated at 68.3% and 95.4% c. l. for the constant and epoch - dependent absorptions , respectively . without a spatially flat universe
prior , we also investigate the @xmath6cdm with the corrected distance modulus of union2.1 sn ia by taking the opacity parameter @xmath31 in 68.3% c. l. range constrained from the previous model - independent tests into consideration .
the linear parametrization for the cosmic opacity is also considered .
the results projecting to the @xmath70 plane are shown in figure [ fig5 ] . because of the marginalization " of @xmath31 , which somewhat weakens the constraints on parameters @xmath26 and @xmath25 , the statement that the expansion of universe is accelerating is less eloquent than the one in ref .
@xcite , which concludes that a currently accelerating universe is needed at 99.9% ( 3.9@xmath71 ) c. l. to agree with their distance measurements of sne ia . however , we find that a decelerating universe with @xmath7 is only allowed at @xmath72 c. l. by observations when the spherical @xmath5 model is taken into account .
so , the standard cosmological scenario is still supported by observations , although current data may favor a universe with non - zero opacity ,
in this paper , by considering the luminosity distances provided by the largest union2.1 sn ia sample together with the add given by galaxy clusters samples , we first examine the possible cosmic opacity in a cosmological - model - independent way .
two redshift - dependent parametric expressions : @xmath2 and @xmath3 are considered to describe the optical depth associated to the cosmic absorption .
the results suggest that the tests of cosmic opacity are not significantly sensitive to the parametrization for @xmath65 . for the add sample compiled by filippis _
et al . _
@xcite with an elliptical @xmath5 model , we obtain that a universe with little opacity ( almost transparent ) is favored . for the add samples given by bonamente _
et al . _
@xcite , where three different cluster gas distribution models are applied , the test results suggest that a transparent universe can only be marginally consistent with observations at 95.4% c. l. ( even at 99.7% c. l. as the @xmath57 kpc - cut spherical @xmath5 model is applied ) .
these are fairly different from the conclusions in refs .
@xcite . by considering the possible cosmic opacity ( 68.3% c. l. range ) constrained from the previous model - independent tests , we obtain the corrected distance modulus of sne ia and then use them to investigate its cosmological implications . in the context of a flat @xmath6cdm ,
the likelihood functions of @xmath26 are examined .
the results are shown in figures ( [ fig3 ] , [ fig4 ] ) .
we find that the opacity parameter @xmath31 constrained from previous cosmological - model - independent tests have slight influences on the probability distributions of @xmath26 and a universe with @xmath69 is required to account for the dimming of the sne ia .
discarding the condition of the spatial flatness , we display the corresponding plots in the @xmath70 plane in figure [ fig5 ] .
we find that a decelerating expanding universe with @xmath7 is only accommodated by observations at @xmath73 c. l. when the spherical @xmath5 model is considered .
that is , a positive cosmological constant is still needed to account for the dimming of sne ia and the standard cosmological scenario remains being supported by current observations .
finally , it should be pointed out that , the presence of systematic uncertainties in observations , especially add measurements using sze+x - ray surface brightness observations , and the assumption of the frequency independency of absorption in the observed frequency range in our analysis might result in some biases of our test results .
moreover , as for the dd relation test , the morphological models of galaxy cluster may also exert a remarkable influence on the tests for cosmic opacity .
in fact , any conclusion that current data may favor a non - zero opacity should be backed up with a thorough analysis of these systematics .
therefore , we may expect more vigorous and convincing constraints on the cosmic opacity within the coming years with more precise data , especially the add data , and a deeper understanding for the absorption in various wavelength bands and the intrinsic three dimensional shape of clusters of galaxies .
we would like to thank a. avgoustidis for helpful discussions .
this work was supported by the national natural science foundation of china under grants nos .
10935013 , 11175093 , 11075083 and 11222545 , the ministry of science and technology national basic science program ( project 973 ) under grant no.2012cb821804 , zhejiang provincial natural science foundation of china under grants nos .
z6100077 and r6110518 , the fanedd under grant no . 200922 , the national basic research program of china under grant no .
2010cb832803 , the ncet under grant no .
09 - 0144 , the pcsirt under grant no .
irt0964 , the hunan provincial natural science foundation of china under grant no .
11jj7001 , and the srfdp under grant no.20124306110001 .
zl was partially supported by china postdoc grant no .2013m530541 .
99 a. g. riess _
et al . _ , aj , 116 , 1009 ( 1998 ) . s. perlmutter _
et al . _ ,
apj , 517 , 565 ( 1999 ) .
a. aguirre , apj 525 , 583 ( 1999 ) .
c. csaki , n. kaloper and j. terning , phys .
88 , 161302 ( 2002 ) .
r. c. tolman , proc .
, 16 , 511 ( 1930 ) . i. m. h. etherington , phil . mag . , 15 , 761 ( 1933 ) . i. m. h. etherington , gen .
39 , 1055 ( 2007 ) .
g. f. r. ellis , gen .
grav . , 39 , 1047 ( 2007 ) .
p. schneider , j. ehlers and e. e. falco , gravitational lenses ( new york : springer , 1999 ) .
j. v. cunha , l. marassi and j. a. s. lima , mnras , 379 , l1 ( 2007 ) .
_ , mnras , 406 , 1773 ( 2010 ) .
e. komatsu _ et al .
_ , apjs , 192 , 18 ( 2011 ) . b. a. bassett and m. kunz , phys . rev .
d 69 , 101 , 305 ( 2004 ) .
b. a. bassett and m. kunz , apj , 607 , 661 ( 2004 ) .
j. p. uzan , n. aghanim and y. mellier , phys . rev .
d 70 , 083553 ( 2004 ) . f. de bernardis , e. giusarma and a. melchiorri , int . j. mod
d 15 , 759 ( 2006 ) .
r. f. l. holanda , j. a. s. lima and m. b. ribeiro , apj , 722 , l233 ( 2010 ) .
r. f. l. holanda , j. a. s. lima and m. b. ribeiro , a&a , 528 , l14 ( 2011 ) .
z. li , p. wu and h. yu , apj , 729 , l14 ( 2011 ) .
r. nair , s. jhingan , and d. jain , jcap05(2011)023 .
r. amanullah _
_ , apj , 716 , 712 ( 2010 ) . e. de filippis , m. sereno , w. bautz and g. longo , apj , 625 , 108 ( 2005 ) . m. bonamente _
_ , apj , 647 , 25 ( 2006 ) .
x. meng , t. zhang , h. zhan and x. wang , apj , 745 , 98 ( 2012 ) .
n. liang , s. cao , k. liao and z. zhu , arxiv:1104.2497 . v.
f. cardone , s. spiro , i. hook and r. scaramella , phys .
d , 85 , 123510 ( 2012 ) .
r. f. l. holanda , j. a. s. lima and m. b. ribeiro , a&a , 538 , a131 ( 2011 ) .
r. f. l. holanda , r. s. goncalves and j. s. alcaniz , arxiv:1201.2378 .
r. s. goncalves , r. f. l. holanda and j. s. alcaniz , mnras , 420 , l43 ( 2012 ) . b. chen and r. kantowski , phys .
d 79 , 104007 ( 2009 ) .
b. chen and r. kantowski , phys .
d 80 , 044019 ( 2009 ) . s. more , t. bovy and d. w. hogg , apj , 696 , 1727 ( 2009 ) .
t. m. davis _
apj , 666 , 716 ( 2007 ) . w. j. percival _
et al . _ , mnras , 381 , 1053 ( 2007 ) . j. chen , p. wu , h. yu and z. li , jcap10(2012)029 .
r. nair , s. jhingan , and d. jain , jcap12(2012)028 .
a. avgoustidis , l. verde and r. jimenez , jcap06(2009)012 .
a. avgoustidis _
et al . _ ,
. m. kowalski _
et al . _ ,
apj , 686 , 749 ( 2008 ) .
d. stern , _
et al . _ , jcap02(2010)008 a. g. riess _
et al . _ ,
apj , 699 , 539 ( 2009 ) .
j. a. s. lima , j. v. cunha and v. t. zanchin , apj , 742 , l26 ( 2011 ) .
r. a. sunyaev and ya .
b. zeldovich , comments astrophys .
space phys . , 4 , 173 ( 1972 ) .
a. cavaliere and r. fusco - fermiano , a&a , 667 , 70 ( 1978 ) .
/ union/. n. suzuki _
et al . _ ,
apj , 746 , 85 ( 2012 ) . b. s. mason , s. t. myers and a. c. readhead , apj , 555 , l11 ( 2001 ) .
e. d. reese _
et al . _ ,
apj , 581 , 53 ( 2002 ) .
p. r. bevington and d. k. robinson , _ data reduction and error analysis for the physical sciences _ , ( mcgraw - hill , boston , ma , 2003 ) , 3rd ed .
, chapter 4 . | with assumptions that the violation of the distance - duality relation entirely arises from non - conservation of the photon number and the absorption is frequency independent in the observed frequency range , we perform cosmological - model - independent tests for the cosmic opacity .
the observational data include the largest union2.1 type ia supernova sample , which is taken for observed @xmath0 , and galaxy cluster samples compiled by de filippis _ et al . _ and bonamente _ et al .
_ , which are responsible for providing observed @xmath1 .
two parameterizations , @xmath2 and @xmath3 are adopted for the optical depth associated to the cosmic absorption .
we find that , an almost transparent universe is favored by filippis _
et al .
_ sample but it is only marginally accommodated by bonomente _
et al . _
samples at 95.4% confidence level ( c. l. ) ( even at 99.7% c. l. when the @xmath4-cut spherical @xmath5 model is considered ) . taking the possible cosmic absorption ( in 68.3% c. l. range ) constrained from the model - independent tests into consideration , we correct the distance modulus of sne ia and then use them to study their cosmological implications . the constraints on the @xmath6cdm show that a decelerating expanding universe with @xmath7 is only allowed at 99.7% c. l. by observations when the bonamente _
et al . _
sample is considered .
therefore , our analysis suggests that an accelerated cosmic expansion is still needed to account for the dimming of sne and the standard cosmological scenario remains being supported by current observations . |
1 . comparison of power - balance prediction ( solid curve ) and bessel - function prediction ( dashed curve ) with numerical integration result ( diamonds ) , at fundamental frequency ( @xmath42 ) .
parameters : ( a ) @xmath44 ; ( b ) @xmath45 . | the inverse ac josephson effect involves rf - induced ( shapiro ) steps that cross over the zero - current axis ; the phenomenon is of interest in voltage standard applications .
the standard analysis of the step height in current , which yields the well - known bessel - function dependence on an effective ac drive amplitude , is valid only when the drive frequency is large compared with the junction plasma frequency or the drive amplitude is large compared with the zero - voltage josephson current . using a first - order krylov - bogoliubov power - balance approach we derive an expression for the threshold value of the drive amplitude for zero - crossing steps that is not limited to the large frequency or large amplitude region .
comparison with numerical solutions of the rsj differential equation shows excellent agreement for both fundamental and subharmonic steps .
the power - balance value for the threshold converges to the bessel - function value in the high - frequency limit .
the term ` inverse ac josephson effect ' was coined by levinsen _
et al .
_ @xcite to describe shapiro steps @xcite in the current - voltage characteristic of a small josephson junction that cross over the zero - current axis .
the phenomenon was recognized as being potentially important for voltage - standard applications inasmuch as the elimination of the dc bias current would eliminate one possible pathway for the entry of noise into the system .
it was studied extensively experimentally , analytically , and computationally by a number of authors ( see , _
e.g. _ , kautz @xcite plus references therein ) .
the ` standard ' analysis that emerged from these studies expressed the height in current of phase - locked shapiro steps in terms of bessel functions of an effective drive amplitude ; this analysis was shown to be appropriate when the frequency of the ac drive is large compared with the junction plasma frequency or when the amplitude of the ac drive is large compared with the zero - voltage josephson current .
our analysis , instead , is based on the power - balance formalism developed by krylov and bogoliubov @xcite ; it is not limited to the large frequency or the large amplitude regions .
comparison of the predictions of the power - balance approach with results obtained by direct numerical integration of the model differential equation typically show agreement to at least three significant digits . in the high - frequency limit
they approach those predicted by the bessel - function expressions .
the starting point of our analysis is the usual shunted - junction model of the small josephson junction , subjected to a dc - bias current and an ac - driving current ; the corresponding differential equation , in normalized form , is @xmath0 where @xmath1 and @xmath2 are , respectively , the dc- and ac - driving currents . the corresponding unperturbed equation , _
i.e. _ , with @xmath3 , @xmath1 , and @xmath2 set to zero , has two types of solutions , _ viz .
_ , oscillatory and rotary . since we are interested in non - zero - voltage states of the junction we focus on the latter , which have the form @xmath4 where am ( ) is the jacobian elliptic amplitude function of modulus @xmath5 , and @xmath6 is an arbitrary constant . the instantaneous junction voltage in this case is thus given by @xmath7 we now return to the original model given by eq
( 1 ) .
an obvious physical requirement for the junction voltage to be a stationary , periodic function similar to that given by eq .
( 3 ) is that the average power dissipated , _
i.e. _ , the conductance times the mean - square voltage , be equal to the average power furnished by the drive currents , _
i.e. _ , the mean - value of the drive current times the voltage . assuming that eq .
( 2 ) can be used as a first approximation to the solution of eq .
( 1 ) in the presence of dissipation and drive , we can write , using results from , the time - average power dissipated as @xmath8 where @xmath9 and @xmath10 are , respectively , the complete elliptic integrals of first and second kinds . assuming first the presence of only a dc - bias current , _
i.e. _ , setting @xmath11 , we can write the time - average input power as @xmath12 equating eq . ( 4 ) to eq .
( 5 ) gives an expression for the mccumber branch of the current - voltage characteristic of the junction in the following parametric form @xmath13 @xmath14 in the zero - bias configuration , instead , _
i.e. _ , with @xmath15 and @xmath16 , we can write the input power as @xmath17 which , by expressing the jacobian dn ( ) function in terms of its fourier - series expansion , we can write as @xmath18\ ! > \ , , \
] ] where @xmath19 , in which @xmath20 is the complementary modulus .
from eq .
( 9 ) we see that @xmath21 only if @xmath22 for some integer @xmath23 , and , assuming eq . ( 10 ) to be satisfied , that @xmath24 can vary smoothly from zero to @xmath25 depending on the value of the phase - shift term @xmath6 ( we have @xmath26 when @xmath27 ) .
the threshold value , @xmath28 , of the drive amplitude is the minimum value for which it is possible to satisfy the equation @xmath29 ; this yields the value @xmath30 for @xmath31 , the constant @xmath6 adjusts itself according to the relation @xmath32=\frac{\gamma_{thr}}{\gamma } \ , .\ ] ] finally , in the general case in which both @xmath33 and @xmath34 , the dissipation defined by eq .
( 4 ) is balanced by both dc and ac inputs .
this permits , _
e.g. _ , calculating the minimum bias - current value to which a phase - locked step extends : assuming eq .
( 10 ) to be satisfied , we find @xmath35 which , for @xmath36 [ see eq . ( 11 ) ] , is a negative number .
the ` standard ' bessel function expression for the height in current of a phase - locked shapiro step is @xcite @xmath37 where @xmath38 ( ) is the bessel function of order @xmath39 .
thus , from eqs .
( 6 ) and ( 13 ) , the threshold value of the drive amplitude in this approach is @xmath40 at this point it must be mentioned that the integer @xmath39 in eq . (
14 ) is _ not _ the same as the integer @xmath23 in eq .
( 10 ) : in fact , the bessel function index @xmath39 refers to a _ super_harmonic number whereas the power balance index @xmath23 refers to a _ sub_harmonic number . to be perfectly clear on this point , the corresponding shapiro steps in the current - voltage characteristic of the junction occur at normalized voltages @xmath41 thus , the only place where the two expressions , eq . ( 11 ) and eq . ( 15 ) , can be compared directly is at the fundamental frequency , where @xmath42 .
our results are summarized in fig .
1a , b . in these figures ,
the solid curve is the power - balance threshold , calculated from eq . ( 11 ) , the dashed curve is the bessel - function threshold , calculated from eq . (
15 ) , and the diamonds are values obtained from the direct numerical integration of eq .
( 1 ) . except for the highest point in fig . 1b , where the numerical result begins to diverge slightly from the power - balance prediction
, these two typically agree to at least three significant digits .
a similar agreement between power - balance prediction and numerical experiment is obtained at the second subharmonic ( @xmath43 ) . for higher frequencies
than those shown in fig .
1a , the two predicted threshold values , eq .
( 11 ) and eq .
( 15 ) , converge to a common asymptotic limit . in order to calculate numerically the points indicated by diamonds in fig . 1
, we often found it necessary to ` tune ' fairly precisely the initial conditions used for eq .
( 1 ) .
this fact suggests that , at least in some cases , the basin of attraction of the phase - locked state is rather small , perhaps indeed vanishing , a fact that would have important implications for the stability of the locked state , _
e.g. _ , against thermal fluctuations .
the technique of cell mapping @xcite might be a useful tool for exploring this question .
braiman _ et al .
_ @xcite have extended the bessel function approach to subharmonic frequencies .
in fact , their eq .
( 2 ) has just the form of the fourier - series expansion of the jacobian am ( ) function of our eq .
( 2 ) .
this approach gives expressions for the step height in current corresponding to eq .
( 14 ) for superharmonic steps in terms of sums of products of bessel functions .
the approach , however , clearly becomes rather unwieldy if more than the first few terms in the fourier expansion are employed .
the power - balance approach can also be extended beyond the lowest level of approximation that we have employed here @xcite ; this would presumably permit a description of superharmonic steps , which presently is lacking from our analysis . however , superharmonic steps occur at progressively higher voltages , which implies progressively higher frequencies , where the simple bessel - function expression , eq .
( 15 ) , is known to give a reasonable description of the situation .
thus , the reward to be obtained here might not be worth the effort .
finally , we must offer one _ caveat _ : as is apparent from fig .
1a , the power - balance prediction becomes progressively better than the bessel - function prediction as the ac drive frequency is reduced below unity ( which is the plasma frequency , with our normalization ) .
this , however , is a region where , in addition to simple periodic solutions of eq .
( 1 ) , exemplified by the ansatz of eq . ( 2 ) , there are known to exist also complicated quasi - periodic and chaotic trajectories @xcite . consequently , whereas eq .
( 11 ) does give a good estimate of the threshold value of the drive amplitude _ if _ a simple step exists , it does _ not _ guarantee the existence of such a step .
we wish to thank roberto monaco for a critical reading of the manuscript and niels grnbech - jensen for illuminating discussions and technical assistance .
b.a.m .
thanks the physics department of the university of salerno for hospitality during the visit that originated this work . financial support from the ec under contract no .
sc1-ct91 - 0760 ( tsts ) of the `` science '' program and from the human capital and mobility " program , from murst ( italy ) , and from the progetto fianalizzato `` tecnologie superconduttive e criogeniche '' del cnr ( italy ) is gratefully acknowledged .
99 m. t. levinsen , r. y. chiao , m. j. feldman , and b. a. tucker , appl .
. lett .
* 31 * ( 1977 ) 776 .
s. shapiro , phys .
rev .
lett .
* 11 * ( 1963 ) 80 .
r. l. kautz , j. appl .
phys . * 52 * ( 1981 ) 3528 .
see , _
e.g. _ , a. scott , active and nonlinear propagation in electronics ( wiley - interscience , new york , 1970 ) , chap .
iii .
p. f. byrd and m. d. friedman , handbook of elliptic integrals for engineers and scientists , second edition ( springer - verlag , berlin , 1971 ) .
y. braiman , e. ben - jacob , and y. imry , in : squid 80 , h. d. halhbohm and h. lbbig , eds .
( de gruyter , berlin , 1980 ) , pp .
783795 .
m. p. soerensen , a. davidson , n. f. pedersen , and s. pagano , phys .
rev .
a * 38 * ( 1988 ) 5384 .
see , _
e.g. _ , n. minorsky , nonlinear oscillations ( van nostrand , new york , 1962 ) , chap . 15 .
r. l. kautz and r. monaco , j. appl .
phys .
* 57 * ( 1985 ) 875 . |
recently a novel idea to bypass this difficulty has been proposed @xcite . by freezing , or pinning , a fraction of the degrees of freedom of the system
, the ideal glass transition temperature has been predicted to rise to a point at which experiments and simulations _ in equilibrium _ are feasible thus allowing to probe the nature of this transition . in ref .
@xcite the authors have studied the effect of pinning for the case of a mean - field spin glass model which is known to exhibit a dynamical mct - transition at a temperature @xmath7 and a thermodynamic transition at a lower temperature @xmath5 .
it was demonstrated that , by pinning a fraction @xmath0 of the degree of freedoms of the spins ( selected at random ) in the equilibrated system , both @xmath1 and @xmath4 increase with @xmath0 .
thus by equilibrating the non - pinned system at an intermediate temperature and subsequently increasing @xmath0 , one can access and probe the ideal glass state in _ thermal equilibrium _
( note that changing @xmath0 does not perturb this equilibrium @xcite ) .
it was found that at sufficiently large @xmath0 the two lines @xmath1 and @xmath4 merge and terminate at a critical point with a universality class of the random - field ising model @xcite . very recently , it has been tested whether this approach to detect @xmath5 in mean field models can also be used in realistic glass formers in finite dimensions @xcite .
it was found that pinned systems do indeed show a behavior that agrees qualitatively with the theoretical predictions , thus giving encouraging evidence that the nature of the ideal glass transition can be studied in equilibrium . in these numerical studies ,
the overlap @xmath8 and its distribution @xmath9 $ ] have been used to identify the amorphous - fluid phase diagram in the @xmath10-@xmath0 plane for pinned systems @xcite . despite these results ,
it is not clear if the so obtained amorphous state is the _ bona fide _ ideal glass , for which the configurational entropy @xmath6 vanishes and , in view of the conceptual importance of @xmath6 , this is a very disturbing situation .
we recall that @xmath6 is related to the number of available states of the system @xcite and is also a key quantity that controls the slow dynamics in the activation regime in which the relaxation time @xmath11 is related to @xmath12 via the adam - gibbs relation , @xmath13 @xcite .
finally , @xmath6 is intimately related to the landau - like free energy associated to the overlap of two coupled replicas @xcite .
although from the simulations reported in ref .
@xcite , the existence of @xmath5 has been inferred from the behavior of the overlap distribution @xmath9 $ ] ( discussed in more detail below ) , such an analysis of the overlap @xmath8 alone may not be conclusive to demonstrate the existence of the arrested phase , since one can not exclude the possibility that other scenarios , such as the purely kinetic one , can also explain the observed features of @xmath9 $ ] .
thus , evaluating @xmath6 directly and identifying the temperature at which it vanishes is crucial to disentangle conflicting theoretical scenarios . in the present work ,
we use computer simulation to determine for a canonical glass - former with pinned particles the ideal glass transition temperature @xmath1 as a point at which @xmath6 vanishes . for the first time
this is done without invoking any kind of extrapolation .
we also calculate the overlap distribution @xmath9 $ ] and find that @xmath5 obtained from @xmath14 and from @xmath9 $ ] agree very well up to a finite value of @xmath0 .
furthermore , we analyze the geometrical properties of the potential energy landscape ( pel ) and use it to evaluate the dynamic transition temperature @xmath4 as a point at which the saddles of the energy landscape vanish .
we find that @xmath4 merges with @xmath1 exactly at the point at which the two mentioned @xmath1 depart from each other , strongly indicating the existence of a critical point which is predicted by the mean - field analysis @xcite .
we study a standard glass - forming model : a three dimensional binary lennard - jones mixture @xcite .
the number of particles is @xmath15 and 300 , but most of the results are for @xmath16 ( see _ materials and methods _ for details ) . throughout the present study ,
the system has been prepared at each temperature by randomly choosing a fraction @xmath0 of particles from the thermally equilibrated samples and quenching their positions ( see _ materials and methods _ for details ) . to obtain the entropy of the pinned system , @xmath17 , we used thermodynamic integration to determine the entropy of a given configuration of pinned particles and subsequently calculated @xmath17 by averaging over the realizations of pinned particles ( see _ materials and methods _ for details ) .
figure 1(a ) shows the entropy per ( unpinned ) particle @xmath18 as a function of the fraction of pinned particles at several temperatures @xmath10 and we recognize that with increasing @xmath0 the entropy decreases rapidly . for all temperatures
this decrease is linear at small @xmath0 but then the curves bend at intermediate @xmath0 and follow a weaker @xmath0-dependence . for @xmath19
this bent becomes sharp , strongly indicating that a _ thermodynamic _ glass transition takes place .
, as evaluated from the thermodynamic integration , as a function of @xmath0 ( symbols ) .
the entropies of the disordered solid states @xmath20 , obtained using the harmonic approximation , are drawn as solid lines .
the dashed lines are a linear extrapolation from the low @xmath0 sides .
( b ) : the configurational entropy @xmath21 .
the error bars have been estimated from the sample to sample fluctuations .
, scaledwidth=48.0% ] [ fig1 ] this becomes more evident by evaluating the configurational entropy obtained by subtracting from @xmath17 the vibrational entropy @xmath22 . in order to estimate @xmath22 ,
we have determined the inherent structures @xcite and calculated the eigenfrequencies @xmath23 . using the harmonic approximation , one can then approximate @xmath24 .
@xmath25 is shown in fig .
1(a ) as well ( solid lines ) and we see that it shows basically a linear decrease with @xmath0 , a trend which is due to the suppression of the low frequency modes in the density of states
. we can now estimate the configurational entropy @xmath26 as the difference @xmath27
@xcite and in fig .
1(b ) we show the @xmath0-dependence of @xmath28 for various temperatures .
this figure shows that , for @xmath29 , @xmath30 quickly decreases with increasing @xmath0 and becomes basically zero _ at a finite value of @xmath0 _ , indicating that the system has entered the ideal glass state in which the entropy is basically due to harmonic vibrations @xcite . for @xmath31 , the approach of @xmath30 to zero
is milder and the bent is less sharp , indicating that the transition becomes a crossover .
we define the ideal glass transition point @xmath32 , or @xmath1 , as the point at which @xmath30 becomes zero . as the temperature
is lowered , @xmath32 decreases and in fig .
4 we show the resulting phase diagram the details of which will be discussed below .
finally we mention that the presented results are for @xmath16 .
however , we have also simulated systems with @xmath15 and found that the results do not depend significantly on @xmath33 ( see si for details ) . an alternative method to locate and characterize
the thermodynamic transition is to study the overlap @xmath34 between two configurations @xmath35 and @xmath36 : @xmath37 , where @xmath38 is the heaviside function , @xmath39 is the position of particle @xmath40 in configuration @xmath35 , and the length - scale @xmath41 is 0.3 @xcite .
rfot predicts that at the glass transition the average value @xmath42 $ ] will increase quickly from a small value in the fluid phase to a large value in the glass phase @xcite .
here @xmath43 and @xmath44 $ ] stand for the thermal and disorder averages , respectively .
we have computed the overlap distribution @xmath45 using replica exchange molecular dynamics @xcite ( see si ) and in fig . 2 we present @xmath9 $ ] for @xmath460.7 and @xmath47 .
for @xmath48 , @xmath9 $ ] remains single - peaked for all @xmath0 , and the peak position shifts continuously towards larger @xmath8 as @xmath0 increases , see fig .
a qualitatively different behavior is observed at @xmath49 ( fig .
2(b ) ) : @xmath9 $ ] is single - peaked at low and high @xmath0 , but has a double peak structure at intermediate @xmath0 , thus signalling the coexistence of the fluid and glass phase , which indicates that the transition from the fluid phase to the glass phase is first - order - like @xcite .
the @xmath0-dependence of the average overlap @xmath42 $ ] is shown in fig .
2(c ) . for high temperatures @xmath50
, @xmath42 $ ] smoothly increases with @xmath0 , reflecting the continuous shift of the single peak of @xmath9 $ ] as shown in fig .
2(a ) . for @xmath19 ,
@xmath51 $ ] shows a quick increase at intermediate values of @xmath0 , in agreement with the presence of the double peak structure seen in @xmath9 $ ] at these @xmath10 .
it suggests that a first - order - like transition , rounded by finite size effect , takes place , in qualitative agreement with the result for a system of harmonic spheres @xcite .
note that , within the accuracy of our data , we see that the amplitude of the ( smeared out ) jump in @xmath42 $ ] seems to vanish around @xmath52 , thus indicating that at around that temperature the line of first order transition ends in a critical point that is second - order - like . from this approach with the overlap
, we can define the ideal glass transition temperature @xmath53 as the temperature at which the skewness of @xmath9 $ ] vanishes ( see si ) , and in fig .
4 we have included the resulting @xmath53 as well . for small and intermediate @xmath0 , we find a very good agreement between @xmath1 and @xmath54 , thus showing that the two very different approaches do give the same ideal glass transition temperature .
this is thus very strong evidence that at this temperature the system does indeed undergo a thermodynamic phase transition from the fluid to an ideal glass state . for temperatures above the second - order - like critical point the curves @xmath1 and @xmath55 differ .
we will rationalize this in the discussion below .
finally we mention that the curves @xmath1 and @xmath53 seem to extrapolate in a smooth manner to the kauzmann temperature of the bulk which has been _ estimated _ from extrapolation @xcite .
this shows that the present measurement of the bulk @xmath5 is compatible with the one from previous estimates .
$ ] for @xmath48 ( a ) and @xmath49 ( b ) .
( c ) : the average overlap @xmath56 $ ] obtained from @xmath9 $ ] as a function of @xmath0 for several temperatures .
the error bars have been calculated using the jackknife method .
, scaledwidth=46.0% ] [ fig - overlap ] in the past , it has been found that the slow dynamics of glass - forming systems is closely related to the features of the potential energy landscape ( pel ) @xcite and in the following we will use these relations to characterize the relaxation dynamics of the pinned system .
figure 3(a ) shows the @xmath2dependence of the average inherent structure energy @xmath57 $ ] . for the bulk system , @xmath58 , @xmath59 $ ]
is basically constant at high temperatures but then steeply decreases below a crossover temperature @xmath60 , a temperature which signals that the relaxation dynamics becomes strongly influenced by the pel @xcite . as @xmath0 increases
, the value of @xmath57 $ ] at high @xmath10 moves steadily upward , which is reasonable since the energy of the system is literally _ pinned _ at the higher energy levels due to the presence of the pinned particles . concomitantly the crossover temperature increases with @xmath0 and
the crossover becomes smeared out , completely disappearing at the highest @xmath0 .
the vanishing of this crossover with growing @xmath0 indicates thus that the pinning affects qualitatively the nature of the pel and of the relaxation dynamics .
for instance it is found that with increasing @xmath0 the fragility of the system decreases and shows at high @xmath0 an arrhenius dependence @xcite . in the inset of figure 3(a ) , the low temperature behavior of @xmath59 $ ] is shown .
it clearly demonstrates that @xmath59 $ ] is inversely proportional to @xmath10 for all @xmath0 . according to the energy landscape scenario ,
this is an indicator that the distribution of @xmath61 is gaussian @xcite .
another important quantity that connects the glassy dynamics of a system with its pel is the saddle index @xmath62 , _
i.e. _ , the number of negative eigenvalues of the hessian matrix at a stationary point of the pel . for bulk systems
it has been found that @xmath62 shows a linear dependence on @xmath63 , the bare energy of a saddle point @xcite .
since the value of @xmath63 at which @xmath62 goes to zero ( this value is often denoted as `` threshold energy '' @xmath64 ) corresponds to the average energy of the inherent structures @xmath65 at the critical temperature of mode - coupling theory , one can extract from the geometrical properties of the pel the value of @xmath7 without having to do any fit to dynamical data @xcite .
we use a standard method to determine numerically the energy and index of saddles for the pinned system ( see si for details ) , and in fig .
3(b ) we plot the average normalized saddle index @xmath66 as a function of its corresponding @xmath63 .
in agreement with previous studies of the pel , we find that @xmath67 decreases linearly as a function of @xmath63 and hence we can obtain @xmath68 from a linear fit ( included in the figure as well ) .
we see immediately that @xmath64 increases with @xmath0 and together with the data of fig .
3(a ) and @xmath69(t_d(c))$ ] , we can conclude that @xmath4 increases with @xmath0 .
the resulting @xmath3dependence of @xmath7 is included in fig .
4 as well and will be discussed in the next section .
we have also evaluated @xmath7 by calculating the relaxation time @xmath11 from the time dependent density correlation function ( see _ material and methods _ ) and by fitting @xmath11 with the mct power - law @xmath70 .
we find that the so obtained values of @xmath71 agree well with those obtained from the pel ( see si for details ) .
note that one needs several fit parameters to obtain @xmath71 from the dynamic data , whereas basically no fit parameter are required for @xmath7 from the pel . in fig . 4 ,
the iso-@xmath11 lines are also plotted .
the graph shows that @xmath11 quickly increases with @xmath0 and that the lines asymptotically approach the @xmath5 line from the high @xmath10-side .
dependence of the averaged inherent structure @xmath57 $ ] for several values of @xmath0 .
inset : same quantity as a function of @xmath72 .
( b ) : the average normalize saddle index , @xmath67 , as a function of its energy , @xmath63 , for several values of @xmath0 .
the threshold energy , @xmath64 , is defined by a linear extrapolation from the high energy side.,scaledwidth=48.0% ] [ fig - saddles ]
in fig . 4 we summarize the results of the previous sections in the form of a phase diagram in the @xmath0-@xmath10 plane .
the ideal glass transition lines @xmath1 determined from @xmath73 and @xmath55 obtained from @xmath45 are plotted as filled circles and diamonds , respectively . at low @xmath10 and
@xmath0 the two temperatures basically coincide but around @xmath74 , they start to depart from each other in that @xmath55 increases continuously whereas @xmath1 bends and its @xmath0-dependence becomes weaker . note that this separation occurs at the same point at which @xmath9 $ ] changes from the bimodal to the single - peaked shape ( see fig .
2 ) , thus indicating that in this region of parameter space the system has a second order like critical point .
the theoretical calculations for a mean - field spin glass model show that @xmath1 should terminate at a finite @xmath0 and that this end point is a critical point of the universality class of the random field ising model @xcite .
furthermore the theory predicts that at this end point the coexistence line @xmath1 and the dynamical line @xmath4 merge .
figure 4 shows that this prediction is indeed compatible with our data in that the two lines do cross near @xmath75 . from the figure we also recognize that , beyond the endpoint , @xmath1 obtained from @xmath73 almost matches with @xmath4
this result is reasonable since in this range of @xmath10 and @xmath0 the particles are strongly confined by the labyrinthine structure imposed by the pinned particles , _
i.e. _ , the system resides mainly at the bottom of a free energy minimum where both the configurational entropy and the number of the saddles vanish simultaneously .
in contrast to @xmath1 , @xmath55 raises continuously even beyond the endpoint .
this can be understood by recalling that this line is defined by the points at which the skewness becomes zero , _
i.e. _ , the point at which @xmath9 $ ] changes from being left - skewed to right - skewed . since this change of sign in the skewness occurs also in the region of the phase diagram beyond the critical point
, the line @xmath54 will extend into that region , similar to the widom line present in a standard liquid - gas transition @xcite . as it is evident from fig .
1(a ) , @xmath5 defined as the point at which @xmath76 becomes somewhat ill - defined beyond the endpoint ( @xmath31 in fig .
1(a ) ) , since the crossover from the fluid to the glass phase becomes broad .
in order to estimate the effect of this ambiguity , we have included in fig . 4 also @xmath77 defined by the linear extrapolation @xmath6 of fig .
1(b ) from the _ low _ @xmath0 side ( open circles ) .
we see that below the end point @xmath19 , @xmath77 is indistinguishable from @xmath5 whereas beyond the end point @xmath50 they bifurcate and @xmath77 becomes comparable with @xmath54 .
at which the configurational entropy vanishes .
the diamonds show @xmath55 determined from the skewness of @xmath9 $ ] .
the squares are dynamic transition points @xmath4 .
the open circles show the ideal glass line @xmath77 determined by the linear extrapolation of @xmath78 to vanish from fig .
the iso - relaxation - times are drawn by @xmath79 , @xmath80 , and @xmath81 .
the star denoted at @xmath58 is a putative ideal glass transition point @xmath82 for the bulk reported in ref .
@xcite . , scaledwidth=48.0% ] [ fig - phasediagram ] to the best of our knowledge , the present study is the first report of a system in finite dimensions that shows the existence of an ideal glass state _ in equilibrium _ , _
i.e. _ , a state in which the configurational entropy is zero at a finite @xmath10 .
the kauzmann temperatures reported in the past have all relied on somewhat questionable extrapolation procedures , leaving thus room for debate over the very existence of a thermodynamic transition @xcite .
our findings are inconsistent with recent simulation studies in which the @xmath2and @xmath3dependence of the relaxation dynamics has been studied @xcite . in ref .
@xcite , the structural relaxation time @xmath11 was fitted with the vogel - fulcher relation @xmath83 $ ] , with @xmath84 as fit parameter .
this relation can be directly derived from the adam - gibbs as well as the rfot theory , both of which assert that @xmath85 , and thus the vogel - fulcher temperature @xmath84 is predicted to be identical to @xmath5 .
furthermore the authors of ref .
@xcite fitted their data also with the mct power - law @xmath86 in order to determine the @xmath3dependence of @xmath7 .
it was found that while @xmath87 increases moderately with @xmath0 , @xmath84 remains constant .
we have plotted the relaxation time @xmath11 as a function of @xmath6 for finite @xmath0 and found that the adam - gibbs relation is violated ( see si for details ) .
thus we conclude that in the case of pinned systems one can not deduce the vogel - fulcher law from the adam - gibbs relation . since the results presented here are all obtained in thermodynamic equilibrium without referring to any kind of extrapolation , we are confident that the phase diagram presented in fig
. 4 does indeed reflect the properties of the system and is not an artifact of the analysis .
a further evidence that the simulated system is really in equilibrium is the observation that the entropy obtained by thermodynamic integration from the high temperature limit matches with that obtained from the low temperature side ( via harmonic approximation ) in the glass phase .
it is also reassuring that all three methods , the thermodynamic integration ( vanishing entropy ) , the overlap distribution ( the discontinuous jump of @xmath8 ) , and the geometric change of the pel , consistently point to the same end point , thus giving strong evidence that this point really exists .
also suggestive is that each combination of pairs amongst three methods are compatible beyond the end point , which is reminiscent of the widom line in the standard gas - liquid phase transition . at this stage
we can conclude that the phase diagram as predicted by the rfot theory is confirmed at least qualitatively .
what remains to be done is to probe the relaxation dynamics in the vicinity of the critical end point since one can expect that this dynamics is rather unusual @xcite and to establish its universality class @xcite .
furthermore , it will also be important to see whether the predicted phase diagram can also be observed in real experiments .
although this will be not easy , for certain systems such as colloids or granular media it should be possible .
the system we use is a binary mixture of lennard - jones particles @xcite . both species @xmath88 and @xmath89 have the same mass and the composition ratio is @xmath90 .
the interaction potential between two particles is given by @xmath91 , where @xmath92 .
we set @xmath93 and @xmath94 .
@xmath95 is truncated and shifted at @xmath96 .
we show energy in units of @xmath97 , with the boltzmann constant @xmath98 , and length in units of @xmath99 .
time units are defined by monte carlo sweeps ( see below ) .
simulations are performed at constant density @xmath100 .
the number of particles is @xmath15 and 300 , and most of the results in the present study are for @xmath16 .
the configuration of the pinned particles is generated by making first a replica exchange run for the bulk system , _
i.e. _ , @xmath58 , using 8 replicas @xcite .
this allows us to generate relatively quickly many equilibrium configurations that are completely independent , _
i.e. _ , between consecutive configurations the mean squared displacement of a tagged particle is more than 100 .
next we use a `` template '' to identify the @xmath101 particles that will be permanently pinned .
details on how to create the template can be found in ref .
this approach has the advantage that the pinned particles cover the space in a relatively uniform manner , thus avoiding the creation of dense regions or large empty regions and hence suppressing strong sample - to - sample fluctuations of the thermodynamic properties of the system .
- thermodynamics : in order to sample thermodynamic properties efficiently at low @xmath10 and large @xmath0 region , we use the replica exchange method @xcite .
the maximum number of replicas is 24 .
more detail is presented in the supplemental information and in ref .
the total cpu time to obtain the presented results is about 580 years of single core time .
+ - dynamics : we use the monte carlo ( mc ) dynamics simulation to calculate dynamical observables @xcite .
the rule of the mc dynamics is the following : in an elementary move , one of the @xmath102 unpinned particle is chosen at random
. then the particle is displaced at random within a cubic box of linear size @xmath103 and the standard metropolis rule is used to decide whether or not the move is accepted .
one mc step consists of @xmath102 such attempts and we set this as a unit of time scale .
the relaxation time @xmath104 is determined by @xmath105 , where @xmath106 is the self part of the intermediate scattering function of the free particles of species @xmath88 for the wave - vector @xmath67 at the peak of the corresponding structure factor .
we have averaged over 30 different realizations of pinned particles to calculate @xmath107 . in order to calculate the entropy @xmath108 of the pinned system
, we have first determined the entropy of the system for a given configuration of pinned particles , and then taken the average over the realization of the configuration of pinned particles . for this
we have calculated the potential energy at temperatures ranging from the target temperature up to the ideal gas limit at @xmath109 , while keeping the pinning configuration fixed .
we have evaluated the entropy of the system with that pinning configuration using thermodynamic integration .
for this integration , we have used a grid in the inverse temperature @xmath110 of width @xmath111 that ranged between 0.01 to 0.1 , depending on the temperature , and integrated the potential energy as a function of @xmath36 .
special care was taken in the very high temperature regime , in order to accurately and rapidly achieve the convergence to high temperature ideal gas limit .
the high temperature expansion of the potential energy of lennard - jones fluid can be written as @xmath112 @xcite .
we have used simulations at very high temperatures to determine the coefficients @xmath88 , @xmath89 , and @xmath113 , and carried out then the thermodynamic integration analytically . to locate the saddles of the potential energy landscape of the system , we have made a minimization of the squared gradient potential @xmath114 used to calculate the inherent structures , @xmath115 includes the position of the pinned as well as unpinned particles , but only the position of the latter are optimized . after having located a saddle with energy @xmath63 , the hessian matrix was diagonalized and we counted the fraction of negative eigenvalues @xmath116 .
the raw data is shown in the si and in fig .
3 ( b ) we present the average index as a function of @xmath63 .
the threshold energy @xmath64 is defined by @xmath117 .
we thank g. biroli , c. cammarota , d. coslovich , and k. kim for helpful discussions .
mo acknowledge the financial support by grant - in - aid for jsps fellows ( 26.1878 ) .
wk acknowledges the institut universitaire de france .
ai acknowledges jsps kakenhi no .
km and mo acknowledge kakenhi no . 24340098 , 25103005 , 25000002 , and the jsps core - to - core program .
the simulations have been done in research center for computational science , okazaki , japan , at the hpc@lr , and the cines ( grant c2014097308 ) .
g. debenedetti and f. h. stillinger , supercooled liquids and the glass transition , _ nature _ , * 410 * , 259 ( 2001 ) .
k. binder and w. kob , _ glassy materials and disordered solids _ ( world scientific , singapore , 2011 ) .
a. cavagna , supercooled liquids for pedestrians , _ phys .
_ , * 476 * , 51 ( 2009 ) .
l. berthier , g. biroli , j .-
bouchaud , l. cipelletti , and w. van saarloos , _ dynamical heterogeneities in glasses , colloids , and granular media _ , oxford university press , oxford , ( 2011 ) .
l. berthier and g. biroli , .
* 83 * , 587 ( 2011 ) .
d. chandler and j. p. garrahan , , 61 , 191 ( 2010 ) .
w. kauzmann , , , * 43 * , 219 ( 1948 ) .
g. tarjus , . in `` dynamical heterogeneities in glasses , colloids , and granular media '' , eds . : l. berthier , g. biroli , j .-
bouchaud , l. cipelletti , and w. van saarloos ( oxford university press , oxford , 2011 ) .
h. tanaka , , , * 23 * , 284115 ( 2011 )
. g. biroli and j .- p .
bouchaud , , in p. g. wolynes and v. lubchenko , eds . , _ structural glasses and supercooled liquids : theory , experiment , and applications _
( wiley , new york , 2012 ) .
t. r. kirkpatrick , d. thirumalai , and p. g. wolynes , , , * 40 * , 1045 ( 1989 ) .
w. gtze , , ( oxford university press , oxford , 2009 ) .
j. kurchan , g. parisi , p. urbani , and f. zamponi , , , * 117 * , 12979 ( 2013 ) .
a. ikeda and k. miyazaki , , 255704 ( 2010 ) . c. cammarota and g. biroli , ideal glass transitions by random pinning , , 109 , 8850 ( 2012 ) .
c. cammarota and g. biroli , random pinning glass transition : hallmarks , mean - field theory and renormalization group analysis , , 138 , 12a547 ( 2013 ) . c. cammarota ,
a general approach to systems with randomly pinned particles : unfolding and clarifying the random pinning glass transition , , 101 , 56001 ( 2013 ) .
l. berthier and w. kob , static point - to - set correlations in glass - forming liquids , , * 85 * , 011102 ( 2012 ) .
p. scheidler , w. kob , and k. binder . .
, * 108 * , 6673 ( 2004 ) .
s. franz and g. parisi , , , p11012 ( 2013 ) .
w. kob and l. berthier , , , 110 , 245702 ( 2013 ) .
g. adam and j. h. gibbs , , , * 43 * , 139 ( 1965 ) . s. franz and g. parisi , . , * 79 * , 2486 ( 1997 ) .
w. kob and h. c. andersen , , , * 51 * , 4626 ( 1995 ) . f. h. stillinger and t. a. weber , , , * 25 * , 978 ( 1982 ) .
f. sciortino , w. kob , and p. tartaglia , , the fact that the difference @xmath118 is slightly negative is related to the harmonic approximation used to estimate @xmath20 .
we have estimated the anharmonic correction to @xmath20 , by considering the nonlinear contribution of the potential energy at the inherent structures @xcite and found that the corrections are very small and _ negative_. evaluation of the exact values of this correction is , however , difficult and beyond the scope of the present work . in principle
one can avoid this difficulty by evaluating the configurational entropy directly @xcite .
however , at present such an approach is computationally way too expensive for systems close to the ideal glass transition .
l. berthier and d. coslovich , , , * 111 * , 11668 ( 2014 ) . f. sciortino , , , p05015 ( 2005 )
. s. sastry , p. g. debenedetti , and f. h. stillinger , , , * 393 * , 554 ( 1998 ) .
y. brumer and d. r. reichman , , , * 69 * , 041202 ( 2004 ) .
k. kim , k. miyazaki , and s. saito , , , * 23 * , 234123 , ( 2011 ) .
g. ruocco , f. sciortino , f. zamponi , c. de michele , and t. scopigno , landscapes and fragilities , , 120 , 10666 ( 2004 ) .
l. angelani , r. di leonardo , g. ruocco , a. scala , and f. sciortino , , , * 85 * , 5356 ( 2000 ) .
k. broderix , k. k. bhattacharya , a. cavagna , a. zippelius , and i. giardina , , , * 85 * , 5360 ( 2000 ) .
b. widom , equation of state in the neighborhood of the critical point , , * 43 * , 3898 ( 1965 ) .
l. xu , p. kumar , s. v. buldyrev , s .- h .
chen , p. h. poole , f. sciortino , and h. e. stanley , relation between the widom line and the dynamic crossover in systems with a liquid
liquid phase transition , , 102 , 16558 ( 2005 ) .
j. p. garrahan , transition in coupled replicas may not imply a finite - temperature ideal glass transition in glass - forming systems , , 89 , 030301 ( 2014 ) .
s. chakrabarty , s. karmakar , and c. dasgupta , , .
s. k. nandi , g. biroli , j .-
bouchaud , k. miyazaki , and d. r. reichman .
critical dynamical heterogeneities close to continuous second - order glass transitions . , * 113 * , 245701 ( 2014 ) .
g. biroli , c. cammarota , g. tarjus , and m. tarzia , random - field - like criticality in glass - forming liquids , , 112 , 175701 ( 2014 ) .
k. hukushima and k. nemoto , exchange monte carlo method and application to spin glass simulations , , 65 , 1604 ( 1996 ) .
l. berthier and w. kob , , , * 19 * , 205130 ( 2007 ) . b. coluzzi , g. parisi , and p. verrocchio , , , * 112 * , 2933 ( 2000 ) .
j. nocedal and s.j .
wright , , ( springer verlag , berlin , 1999 ) . *
supplementary information *
here we describe the details of the replica exchange method which we have employed in this study , which is basically the repetition of the description in ref . @xcite and in its supplemental information .
we have used the parallel tempering algorithm @xcite using up to 24 replicas . in this approach ,
one simulates simultaneously several copies of the system _ i.e. _ , the same hamiltonian but each replica is at a different temperature . using a boltzmann criterion and the detailed balance condition , we periodically attempt to interchange the configurations of two replicas at different temperatures . hence each replica makes a random walk in the temperature space . due to the fast relaxation at high @xmath10
, this will lead to an efficient relaxation of the system also at low @xmath10 .
the smallest difference in temperature between the two neighboring replicas was @xmath119 , which guarantees a good overlap of the two potential energy distributions .
attempts to switch the two neighboring replicas have been made every 50000 time steps .
we have checked that each parallel tempering run has indeed reached equilibrium by following any given replica and making sure that all temperatures have been sampled sufficiently . a typical path of a replica in temperature space is shown in fig .
si-[fig:1](a ) .
the parallel tempering algorithm indeed allows for the system to be equilibrated down to @xmath120 even when the concentration of pinned particles , @xmath0 , is large .
this can be recognized from fig .
si-[fig:1](b ) , where we show the mean squared displacement ( msd ) of the particles ( distinguishing the type @xmath88 and @xmath89 particles ) as a function of time .
the figure shows that at sufficiently long times the msd becomes very large , indicating that the particles do indeed move through the box also at high values of @xmath0 , even if their relative arrangement does not change much , _
i.e. _ , one is in the glass state . to calculate physical quantities ,
we have followed a given replica in the temperature space and considered only the time intervals at which this replica was at the target temperature . , @xmath121 .
( b ) : the mean squared displacement of @xmath88 and @xmath89 particles for @xmath122 , @xmath121 . ]
for our numerical calculation of the entropy , we employed the thermodynamic integration technique . although this technique is well - known and has long been applied to bulk systems @xcite , its extension to pinned systems is not trivial . in this section ,
we summarize the ideas and present the details of our calculation .
we follow the notations introduced in the method and materials section .
our starting system is a binary mixture of type a and b particles .
since we consider a random pinning of this system , our system is now composed of four different types of particles : @xmath123 pinned particles of type a , @xmath124 pinned particles of type b , @xmath125 unpinned particles of type a , and @xmath126 unpinned particles of type b. the total pinned and unpinned particle numbers are @xmath127 and @xmath128 , respectively .
we denote the coordinate of pinned particles by @xmath129 and unpinned particles by @xmath130 .
first we consider the thermodynamics of the system with a given configuration of pinned particles @xmath131 .
in this situation , the partition function is defined as @xmath132 . \end{aligned}\ ] ] and the thermal average of any variable is @xmath133 , \end{aligned}\ ] ] where @xmath134 is the thermal de broglie wavelength , @xmath135 the potential energy , and @xmath136 characterizes the temperature of the system .
the free energy and the entropy are now written as @xmath137 where @xmath138 is the hamiltonian of the unpinned particles . in this work ,
we generate pinned particles configurations by fixing a fraction @xmath0 particles in the equilibrium configurations , and study the thermodynamics after the disorder average over the realizations of pinned particles .
when the pinned particles are generated at a temperature @xmath139 , their distribution function is given by @xmath140,\ ] ] where @xmath141 is the partition function of the bulk system at the temperature @xmath139 .
accordingly the thermodynamic quantities to be calculated are given by @xmath142_{\beta ' } \equiv \int d{\mbox{\boldmath $ s$ } } \
p_{\rm pin}({\mbox{\boldmath $ s$ } } , \beta ' ) { \langle a \rangle}_{{\mbox{\boldmath $ s$}},\beta}.\end{aligned}\ ] ] the subscript @xmath36 of the thermal average @xmath143 indicates that the thermal average for unpinned particles are taken at @xmath36 , while the subscript @xmath139 of the disorder average @xmath144 $ ] indicates that the pinned particles configurations are obtained from the equilibrium configurations at the temperature @xmath139 . because we are interested mainly in the case @xmath145 , the free energy and the entropy of the system
is now written as @xmath146_{\beta } , \ \ \
s(c,\beta ) = [ \tilde{s}({\mbox{\boldmath $ s$}},\beta)]_{\beta}. \end{aligned}\ ] ] note that when @xmath145 , the thermodynamic average of a mechanical variable @xmath88 becomes exactly @xmath142_{\beta } = \frac{1}{z(\beta ) \lambda^{3n } n_a ! n_b ! } \int d{\mbox{\boldmath $ s$ } } d { \mbox{\boldmath $ r$ } } \ a \exp[-\beta u ] , \label{equiv}\end{aligned}\ ] ] which is nothing but the thermodynamic average for the bulk system at the temperature @xmath36 .
this equivalence is the well - known fact for this type of the pinned system @xcite . to calculate the entropy @xmath147 from simulations , we employ the thermodynamic integration ( ti ) method .
the ti method that is most frequently applied to lennard - jones particles is to connect the ideal gas state and the state of interest by a combination of a compression path ( increasing density ) and a cooling path ( decreasing temperature ) @xcite .
however for pinned fluids , it is not clear how compression should be defined .
therefore we use only the cooling path from the ideal gas limit @xmath109 to the target temperature without changing the density .
this version of the ti has been used for bulk lennard - jones particles @xcite , and here we apply it to the pinned system .
we first apply the ti to the entropy of a given pinned particles configuration , and then take the disorder average over realizations .
the entropy of the system with the pinned particles @xmath131 at the target temperature @xmath148 can be expressed as @xmath149 @xmath150 is the entropy of @xmath151 unpinned particles at @xmath152 . since the interactions between the particles including pinned and unpinned particles become irrelevant if @xmath152 , this term is nothing else than the ideal gas entropy of bulk @xmath151 particles .
the thermal average of the potential energy can be decomposed as @xmath153 where @xmath154 , @xmath155 , and @xmath156 are the potential energies of unpinned particles , pinned particles and the interaction between pinned and unpinned particles , respectively .
here we use the fact that @xmath155 is solely dependent on the configuration of pinned particles and free from the thermal average of unpinned particles . plugging these into eq .
( [ ti0 ] ) and taking the disorder average over realizations , we get the final expression @xmath157_{\beta^{\ast } } \\ & - \int^{\beta^{\ast}}_{0 } d\beta \ [ { \langle u_{\rm up } \rangle}_{{\mbox{\boldmath $ s$}},\beta } + { \langle u_{\rm int } \rangle}_{{\mbox{\boldmath $ s$}},\beta}]_{\beta^{\ast}}. \end{aligned } \label{eq : s_final}\ ] ] in the integral , the temperature for the thermal average runs from @xmath158 to @xmath159 , while the temperature for the disorder average over realizations of the pinned particles is fixed to be the target temperature @xmath148 .
thus we can not use the relation eq .
( [ equiv ] ) to evaluate the integrands .
instead we need to directly calculate the potential energy at @xmath160 $ ] under a given pinned particles configuration @xmath131 and take the disorder average over realizations .
the integral in eq .
( [ eq : s_final ] ) is decomposed into three temperature regimes and each of them is evaluated separately .
( 1 ) for low temperature regime , we calculate the potential energy from the configurations generated by the parallel tempering calculations outlined in sec .
i. from each trajectory , we obtain the thermal average of the potential energies @xmath154 and @xmath156 at different temperatures with a given pinned particles configuration .
we employ the simpson s rule to evaluate the integral and take the disorder average over realizations .
( 2 ) above the highest temperature in the parallel tempering calculations , we use standard monte carlo method to calculate the potential energy .
we slice the temperature into the grids with the width @xmath161 depending on the temperature range and calculate the thermal average of @xmath154 and @xmath156 at each temperature for a given pinning configuration .
we employ the simpson s rule to evaluate the integral and take the disorder average over realizations .
( 3 ) we take a special care about the integration at very high temperature @xmath162 , since the potential energies @xmath154 and @xmath163 diverge in the high temperature limit . to accurately calculate the integral
, we first fit the potential energy data to a polynomial function then analytically integrate the function .
considerations on the high temperature expansion shows that the potential energy of the lennard - jones particles at high temperature behaves as @xmath164 where @xmath88 , @xmath89 and @xmath113 are constants @xcite .
we thus fit the data of @xmath165 with expression ( [ eq : u_expansion ] ) to obtain these constants and integrate it analytically . in fig .
si-[fig:2 ] , we show the raw data of @xmath166 and the fitting function eq .
( [ eq : u_expansion ] ) as a function of @xmath167 , confirming that eq .
( [ eq : u_expansion ] ) holds well . as a function of @xmath167 for very high @xmath10 ( @xmath168 ) .
the raw data ( symbols ) are fitted well by the fitting curves ( red solid lines ) .
the vertical dashed line indicates the temperature above which we use the analytical integration .
note that we show here only the data for one realization of the pinned particles since the curves will depend on the realization . ]
we next summarize the method to calculate the harmonic vibrational entropy of the pinned system .
we consider the system that weakly vibrates around an inherent structure ( is ) .
if we denote by @xmath169 the displacement of the @xmath40-th particle from its position in the inherent structure , the potential energy can be approximated well as @xmath170 it is important to realize that only the derivative of the potential energy respect to the coordinates of unpinned particles should be taken into account , not including the ones of pinned particles ( but of course @xmath135 will depend on the positions of the pinned and unpinned particles ) .
thus the size of the hessian matrix is @xmath171 . introducing the eigenvalues @xmath172 of the hessian
, the harmonic vibrational entropy of the given inherent structure with a given pinned particle configuration can be written as @xmath173 note that the eigenvalues @xmath174 depends on the choice of the inherent structure and the pinned particle configuration . taking the average of @xmath175 over realizations of pinned particle configurations and the inherent structures , we finally obtain the harmonic vibrational entropy of pinned system @xmath176 . in practice
, we have sampled the inherent structure by minimizing the potential energy of instantaneous configurations obtained in the simulations . for this calculation we used the conjugate - gradient method .
note that the pinned particles were frozen during the minimization process and only the coordinates of the unpinned particles are optimized . in fig .
si-[fig:3 ] , we show @xmath6 for @xmath15 and 300 for several state points .
we find that finite size effects are small , at least for these two system sizes . and @xmath16 .
we have determined the averaged distribution function @xmath9 $ ] of the overlap @xmath8 as explained in the main text .
subsequently we have calculated the skewness @xmath177 as @xmath178(q-[\langle q\rangle ] ) ^3 } { \left ( \int_0 ^ 1 dq [ p(q ) ] ( q- [ \langle q\rangle])^2\right)^{3/2 } } \quad .\ ] ] the @xmath2and @xmath3dependence of @xmath179 is shown in fig .
si-[fig:4 ] .
the point at which @xmath179 is zero is used to define the kauzmann temperature @xmath54 shown in fig . 4 of the main text . of the distribution function @xmath9 $ ] .
the horizontal dashed line is the zero - axis and used to define @xmath54 .
the error bars have been calculated using the jackknife method . ]
vs @xmath63 . ( b ) : the @xmath57 $ ] as a function of @xmath10 .
the horizontal and vertical arrows indicate the location of @xmath68 and @xmath4 , respectively . this plot is used to map @xmath68 to @xmath4 . ] in fig .
si-[fig:5](a ) , we show the original data of fig
. 3(b ) of the main text .
this scatter plot shows the normalized saddles index @xmath67 as a function of the energy at the saddle , @xmath63 , obtained from the minimization of the square gradient potential latexmath:[$w=\frac{1}{2 } |\nabla u averaging over this scatter data for a given @xmath63 .
the threshold energy @xmath64 is extracted from fig .
3(b ) of the main text as a point at which the averaged @xmath67 vanishes .
mapping of @xmath68 to the dynamic transition temperature @xmath4 can be done by plotting the temperature dependence of the inherent structures @xmath57 $ ] .
the reason why we use the inherent structures @xmath59 $ ] instead of @xmath181 $ ] is that in practice one can evaluate the @xmath2dependence of @xmath57 $ ] with higher precision than that of @xmath181 $ ] .
note that using @xmath182 $ ] would in fact give the same result since it is expected to be very close to @xmath183 $ ] in the low temperature regime . in fig .
si-[fig:5](b ) we show the @xmath10-dependence of @xmath57 $ ] for several @xmath0 s .
the inherent structures are found to be a monotonic function of @xmath10 for all @xmath0 s and therefore @xmath64 can be uniquely mapped to @xmath7 . where @xmath184 is a fitting parameter .
the solid lines indicate the mct power - law .
( b ) : phase diagram including the @xmath184(c ) line . ]
the dynamic transition temperatures @xmath7 can be evaluated by two methods : the dynamic route and the static one @xcite . in the dynamic route , @xmath7 is obtained directly by fitting the relaxation time @xmath11 from the time - dependent correlation functions , such as the scattering function @xmath185 , to the mct power - law @xmath186 .
thus in this case one has @xmath187 , @xmath179 , and the prefactor are fit parameters , all of which will depend on @xmath0 . in the present paper , we have adopted the alternative static route to obtain @xmath7 defined as a point at which the saddles of the potential energy landscape vanish . for bulk systems , @xmath58 , it has been found that @xmath187 from fitting ( dynamic ) agrees very well with @xmath87 as determined from the saddles ( static ) @xcite . .
within the adam - gibbs theory one expects that this representation of the data gives rise to a linear dependence , which for @xmath188 is evidently not the case . ]
we follow exactly the same method to obtain @xmath184 for the pinned systems , i.e. @xmath189 . in fig .
si-6(a ) , we plot @xmath104 obtained from @xmath106 as a function of @xmath190 for several values of @xmath0 . both @xmath184 and @xmath179 are used as fitting parameters to obtain the good fit with the power - law .
as it is well known for bulk systems , special care has to be taken to chose an optimized temperature windows over which @xmath11 can be fitted by a power - law and we have thus put the emphasis to get a good fit in the low temperature part of the data . in fig .
si-6(b ) we plot the so obtained @xmath191 in the phase diagram presented in the main text .
it is clear that the agreement of the two dynamic transition temperatures obtained from different routes is very good .
however , we also recognize that the @xmath0-dependence of @xmath192 is noisier .
this is most likely due to the subtleties of the fitting procedures since three free fit parameters have to be used for @xmath191 ( @xmath184 , @xmath179 , and prefactor ) , whereas no fitting parameter is required to evaluate @xmath7 from the saddles .
this is the reason why in the present study we put more emphasis on the static method .
in this section we will show that for the pinned fluid the adam - gibbs relation @xmath193 @xcite is violated . in fig .
si-7 , we plot the relaxation time @xmath11 against @xmath194 for different values of @xmath0 . for @xmath58 ,
one observes that @xmath195 is indeed a linear function of @xmath194 and that therefore the ag relation holds , as it has been already documented in refs.@xcite .
however , as @xmath0 increases , @xmath196 systematically deviates from the linear dependence and becomes a convex function of @xmath194 .
thus this result clearly demonstrates the violation of the ag relation for pinned fluids . | we use computer simulations to study the thermodynamic properties of a glass former in which a fraction @xmath0 of the particles has been permanently frozen . by thermodynamic integration , we determine the kauzmann , or ideal glass transition , temperature @xmath1 at which the configurational entropy vanishes .
this is done without resorting to any kind of extrapolation , _
i.e. _ , @xmath1 is indeed an equilibrium property of the system .
we also measure the distribution function of the overlap , _
i.e. _ , the order parameter that signals the glass state . we find that the transition line obtained from the overlap coincides with that obtained from the thermodynamic integration , thus showing that the two approaches give the same transition line .
finally we determine the geometrical properties of the potential energy landscape , notably the @xmath2and @xmath3dependence of the saddle index and use these properties to obtain the dynamic transition temperature @xmath4 .
the two temperatures @xmath1 and @xmath4 cross at a finite value of @xmath0 and indicate the point at which the glass transition line ends .
these findings are qualitatively consistent with the scenario proposed by the random first order transition theory . upon cooling ,
glass - forming liquids show a dramatic increase of their viscosities and relaxation times before they eventually fall out of equilibrium at low temperatures @xcite .
this laboratory glass transition is a purely kinetic effect since it occurs at the temperature at which the relaxation time of the system crosses the time scale imposed by the experiment , _
e.g. _ , via the cooling rate . despite the intensive theoretical , numerical , and experimental studies of the last five decades , the mechanism responsible for the slowing down and thus for the ( kinetic ) glass transition is still under debate and hence a topic of intense research . from a fundamental point of view
the ultimate goal of these studies is to find an answer to the big question in the field : does there exist a _
finite _ temperature at which the dynamics truly freezes and , if it does , whether this _ ideal _ glass transition is associated with a thermodynamic singularity or whether it is of kinetic origin @xcite .
support for the existence of a kinetic transition comes from certain lattice gas models with a `` facilitated dynamics '' @xcite . in these models ,
the dynamics is due to the presence of `` defects '' and hence for such systems the freezing is not related to any thermodynamic singularity .
however , the first evidence that there does indeed exist a thermodynamic singularity goes already back to kauzmann who found that the residual entropy ( the difference of the entropy of the liquid state from that of the crystalline state ) vanishes at a finite temperature @xmath5 if it is extrapolated to temperatures below the laboratory glass transition @xcite .
subsequently many theoretical scenarios that invoke the presence of a thermodynamic transition have been proposed @xcite .
one of these is the so - called `` random first order transition '' ( rfot ) theory which , inspired by the exact solution of a mean - field spin glass , predicts that at @xmath5 the glass - former does indeed undergo a thermodynamic transition at which the residual , or configurational entropy @xmath6 ( the logarithm of the number of the states which are available to the system ) vanishes and concomitantly breaks the replica symmetry @xcite .
a further appealing feature of rfot is that it seems to reconcile in a natural way the ( free ) energy - landscape scenario and mode - coupling theory ( mct ) , a highly successful theory that describes the relaxation dynamics at intermediate temperatures @xcite . despite all these advances ,
the arguments put forward in the various papers must be considered as phenomenological since compelling and undisputed experimental or numerical evidence to prove or disprove any of these theories and scenarios is still lacking .
the only exception are hard spheres in infinite dimensions , for which mean - field theory should become exact @xcite , but even in this case some unexpected problems are present , see ref .
@xcite .
this lack of understanding is mainly due to the steep increase of the relaxation times which hampeirs the access to the transition point of thermally equilibrated systems and hence most of the efforts to identifying the transition point , if it exists , resort on unreliable extrapolation . |
bubble sort is an elementary ( and inefficient ) sorting algorithm that proceeds in a number of passes . in each
pass the sequence to be sorted is scanned from left to right . in any pass every item found to be greater than the item immediately to its right is exchanged with this item .
as the algorithm proceeds items move to the right ( bubble up the sequence ) until blocked by some larger item . in general , many passes are required before the sorting is complete . in this paper
we are interested in the permutational effect of a single pass of bubble sort .
we shall see that this effect has some unexpected connections with the theory of permutation patterns .
but before stating our results we need to define our terms more precisely .
we shall only consider sequences of distinct terms .
such a sequence is order isomorphic to a unique permutation and the bubble sort algorithm would process this permutation in the same way as it processes the original sequence .
we state many of our results in the language of permutations ; this gains clarity without losing generality .
all permutations in the paper are on the set of terms @xmath1 for some @xmath2 .
roman letters denote single terms of a sequence and greek letters denote ( possibly empty ) sequences .
notation such as @xmath3 is shorthand for @xmath4 for all terms @xmath5 of @xmath6 .
the operator @xmath0 that describes the effect of a single pass of bubble sort is easily seen to have the following recursive definition . for the empty sequence @xmath7 we have @xmath8 and for non - empty sequences @xmath9 written as @xmath10 where @xmath11 is the maximal term , we have @xmath12 an alternative definition of @xmath0 is furnished by the easily proved [ bdefinition2 ] if @xmath13 where @xmath14 are the left to right maxima of @xmath9 then @xmath15 the other concept we need is that of a pattern class of permutations
. we shall write @xmath16 to denote that @xmath17 is a subsequence of @xmath6 ( although not necessarily a consecutive subsequence ) .
we say that a permutation @xmath9 is a subpermutation of a permutation @xmath18 if @xmath18 has a subsequence that is order isomorphic to @xmath9 , and denote this by @xmath19 .
for example , @xmath20 because @xmath21 .
the subpermutation relation is a partial order and it is studied through its downsets which are called pattern classes .
each pattern class @xmath22 can be characterised by the minimal set @xmath23 of permutations that it avoids : @xmath24 the set @xmath23 is called the basis of @xmath22 and , if @xmath25 , @xmath22 is called a _ principal _ pattern class .
our first connection between the operator @xmath0 and pattern classes is very easy .
@xmath26 is an increasing permutation if and only if @xmath27 .
let @xmath28 be a permutation with largest term @xmath11 .
then @xmath9 is sorted by @xmath0 if and only if @xmath29 is sorted by @xmath0 , @xmath30 is increasing , and @xmath31 .
but , by induction on @xmath32 , this occurs if and only if @xmath29 avoids @xmath33 and @xmath34 , @xmath30 is increasing , and @xmath31 , which is if and only if @xmath9 itself avoids @xmath33 and @xmath34 .
this result , which characterises the permutations sortable by a single pass of bubble sort , can be expressed in another way using the fact that the increasing permutations are precisely those that avoid the permutation @xmath35 : @xmath36 at this point it is convenient to contrast the sorting operator @xmath0 with a similar operator @xmath37 introduced by julian west @xcite .
the definition of @xmath37 on permutations of length @xmath38 is @xmath39 with @xmath40 .
west introduced this operator in the context of sorting via one pass through a stack . here
the permutations that @xmath37 can sort are precisely those of @xmath41 .
the effect of composing the operator @xmath37 with itself has been studied extensively @xcite .
it is natural to ask questions about the compositions of such operators .
one such question is : which permutations can be sorted by applying @xmath0 then @xmath37 ?
in other words what is the set @xmath42 in this note we shall answer a much more general question . we shall determine the permutations @xmath43 for which @xmath44 is a pattern class and , when it is , give its basis .
our results are stated in terms of the number of left to right maxima of a permutation @xmath43 .
we begin with a result that shows that it is rare for @xmath44 to be a pattern class . in this result
( and subsequently ) we write @xmath45 for @xmath46 for typographical convenience .
[ notaclass ] [ lemma : lmax>2 ] if @xmath43 is a permutation with at least three left to right maxima , the third of which is not the final symbol of @xmath43 , then @xmath44 is not a pattern class .
we begin by noting that , if @xmath43 has length @xmath38 and does not end with its maximal element , then @xmath47 .
the reason for this is that , if @xmath48 is a permutation with @xmath49 , then we have @xmath50 where the last equivalence is a consequence of @xmath43 not ending with @xmath38 .
now let @xmath43 be a permutation of length @xmath38 with at least three left to right maxima the third of which is not its final term .
by the first remark of the proof we may ( by appending a new maximal element to @xmath43 if necessary ) assume that @xmath43 ends with its largest term .
thus we may take @xmath43 to have the form @xmath51 where @xmath52 are the first three left to right maxima of @xmath43 and @xmath53 is non - empty .
consider the pair of permutations @xmath54 and @xmath55 .
then , as @xmath56 , @xmath57
. on the other hand , if there were an embedding of @xmath43 into @xmath58 , @xmath59 could not map onto @xmath60 and so , as @xmath61 is only one term longer than @xmath43 , @xmath62 would map onto @xmath63 which is impossible as @xmath64 but @xmath65 .
thus @xmath61 does not contain @xmath43 and so @xmath66 . since @xmath67 we have proved that @xmath44 is not a pattern class .
in the remainder of this section we shall prove a series of results that yield a strong converse of theorem [ notaclass ] .
we shall not only show that @xmath44 is a pattern class for all permutations @xmath43 not covered by theorem [ notaclass ] but will display an explicit basis .
there are several cases to consider but our basic methodology is to identify sets of permutations @xmath68 to which the following lemma can be applied .
[ basicmethod ] let @xmath43 be any permutation .
if there exists a set @xmath68 of permutations such that for any permutation @xmath9 : 1 .
@xmath69 for some @xmath70 , 2 .
@xmath71 for some @xmath70 @xmath72 , then @xmath44 is a pattern class .
furthermore if @xmath68 is a minimal set with these properties then @xmath68 is the basis of @xmath44 .
the two conditions say that @xmath71 for some @xmath70 if and only if @xmath73 .
however @xmath74 in other words @xmath75 for all @xmath70 if and only if @xmath76 .
this shows that @xmath44 is a downset in the subpermutation order , i.e. a pattern class .
it also shows that , if @xmath68 is minimal , it is the basis of @xmath44 .
we first dispose of two trivial cases : [ trivialcases ] if @xmath43 is the permutation of length 1 then @xmath44 is empty . if @xmath77 then @xmath44 consists of the permutation 1 alone .
the first statement is trivial because @xmath78 is empty . in the second case
, if a permutation @xmath9 lies in @xmath79 then @xmath26 is decreasing .
but @xmath26 ends with its maximal term and hence @xmath80 .
we next consider the general case that @xmath43 has a single left to right maximum . to do this we prove the following two lemmas which verify the two conditions of lemma [ basicmethod ] .
[ onelr1 ] let @xmath9 , @xmath81 be sequences both of length greater than 1 such that @xmath81 begins with its largest term and such that @xmath82 .
then there exists @xmath83 such that @xmath9 contains one of @xmath84 and @xmath85 .
we shall prove the result by induction on @xmath32 .
if @xmath86 the result is vacuously true since @xmath82 is possible only if @xmath87 is empty ( as @xmath26 ends with its maximal term ) and this is impossible as @xmath88 .
so now assume that @xmath89 and that the result holds for sequences shorter than @xmath9 .
writing @xmath10 , where @xmath11 is the largest term of @xmath9 , we have @xmath90 .
in fact , as @xmath81 does not end with its largest term , we have @xmath91 .
we consider the various ways in which @xmath81 can lie across @xmath92 .
suppose first that @xmath93 with @xmath94 non - empty and that @xmath95 and @xmath96 .
since @xmath29 is shorter than @xmath9 the inductive hypothesis applies and it proves that , for some @xmath83 , @xmath29 contains one of @xmath97 and @xmath98 .
but then @xmath10 contains one of @xmath99 and @xmath100 .
suppose next that @xmath101 and that @xmath102
. then @xmath103 .
finally suppose that @xmath104
. then @xmath105 .
[ onelr2 ] let @xmath87 be any sequence and @xmath106 values with @xmath107 .
if @xmath9 is a sequence that contains either of @xmath84 or @xmath85 then @xmath81 is contained in @xmath26 .
if @xmath108 then no symbol of @xmath81 can be a left to right maximum of @xmath9 .
however lemma [ bdefinition2 ] implies that @xmath0 preserves the order of symbols that are not left to right maxima ; hence @xmath82 .
this argument would apply to the case @xmath109 if @xmath110 was not a left to right maximum ( obviously the symbols of @xmath87 can not be left to right maxima ) . however
if @xmath110 was a left to right maximum then @xmath5 or some symbol between @xmath110 and @xmath5 must also be a left to right maximum . by lemma [ bdefinition2 ] again this other
left to right maximum precedes @xmath87 in @xmath9 and therefore @xmath110 precedes @xmath87 in @xmath26 .
lemmas [ basicmethod ] , [ onelr1 ] , [ onelr2 ] now handle the case of permutations with just one left to right maximum : [ onelrmain ] suppose that @xmath111 has length greater than @xmath112 and begins with its maximal element .
then @xmath44 is a pattern class with basis @xmath113 [ aalphab ] if @xmath114 has only two left to right maxima , namely @xmath11 and @xmath38 , then @xmath44 is a pattern class with basis @xmath113 note that @xmath115 does not end with its maximum element whereas any permutation @xmath26 does end with its maximum element .
this means that @xmath116 hence @xmath117 thus @xmath118 and the result follows from the previous proposition . before treating the case that @xmath43 has two or three left to right maxima we introduce a variant of the standard diagrammatic way of displaying permutations which will be a helpful aid in understanding the bases of the pattern classes @xmath44 .
every permutation @xmath119 can be represented by its graph of points @xmath120 drawn in the plane . to specify a permutation , only the vertical and horizontal orders of points matter rather than their precise values and so these graphs are useful tools for arguing about subpermutations .
for example the graph of figure [ diagram ] shows the permutation @xmath121 together with a subpermutation . and , in white dots ,
the subpermutation @xmath122.,width=144 ] in these permutation diagrams no two points can be vertically aligned or horizontally aligned since then their positional order or their value order would be ambiguous .
however , if we want diagrams to display a _ set _ of permutations we can exploit this very ambiguity .
for example figure [ permset ] represents a set of 4 permutations because two points lie on the same horizontal line and two lie on the same vertical line .
our major use of this notation is in figure [ basispermextension ] . on the left
this shows a permutation @xmath43 with two left to right maxima ( the upper left quadrant being empty ) . on the right
it shows a set of permutations @xmath123 .
the two white points lie anywhere in the range indicated .
these permutations are extensions of @xmath43 . in all but four cases they are 2-point extensions .
however it is permitted that the white points can coalesce ( in a point in the top left corner ) and then the diagram represents four 1-point extensions of @xmath43 .
this set of permutations ( or , more precisely , the minimal permutations of the set ) will turn out to be the basis of @xmath44 ( except when @xmath6 is empty ) .
and the set @xmath123.,width=384 ] [ twolr1 ] let @xmath9 , @xmath124 be sequences such that @xmath110 and @xmath5 are the only two left to right maxima of @xmath124 with @xmath125 non - empty and such that @xmath126 .
let @xmath43 be the permutation order isomorphic to @xmath124 .
then @xmath9 contains a sequence order isomorphic to one of the members of @xmath123 .
we shall prove that @xmath9 has a subsequence of the form @xmath127 or @xmath128 where 1 .
@xmath93 , 2 .
@xmath129 , 3 .
@xmath130 and @xmath131 are the two largest terms of this sequence , 4 .
if @xmath110 precedes @xmath132 and @xmath94 is empty , then @xmath132 and @xmath130 are the same term .
such a subsequence is indeed isomorphic to a permutation in @xmath123 : the subsequence @xmath133 is order isomorphic to @xmath43 and @xmath134 play the roles of the two white points in figure [ basispermextension ] .
we shall use induction on the length of @xmath9 .
the inductive base is provided by the vacuous cases @xmath135 so we now assume that @xmath136 and that the result is true for shorter sequences ( and any appropriate sequences @xmath124 ) .
writing @xmath10 , where @xmath11 is the largest term of @xmath9 , we have @xmath137 .
in fact , as @xmath124 does not end with its largest term , we have @xmath138 .
we consider the various ways in which @xmath124 can lie across @xmath92 .
@xmath139 and @xmath140 for some partition @xmath141 with @xmath142 non - empty . here
the inductive hypothesis applies and yields a subsequence of @xmath29 of the form @xmath143 or @xmath144 to which the subsequence @xmath145 of @xmath30 may be appended to obtain a subsequence of @xmath9 of the required form .
@xmath146 and @xmath147 . here
@xmath29 can not have length 1 and so lemma [ onelr1 ] shows that @xmath29 has a subsequence @xmath148 or @xmath149 for some @xmath150 .
either @xmath5 lies to the left of @xmath87 in @xmath29 , in which case we can take it to play the roles of both @xmath132 and @xmath130 , or it lies between two terms of @xmath87 or to the right of @xmath87 and , in this case , we take it in the role of @xmath130 and take @xmath151 in the role of @xmath132 .
we can append to this sequence the subsequence @xmath152 of @xmath30 , with @xmath11 playing the role of @xmath131 .
this gives the required subsequence .
@xmath95 and @xmath153 for some partition @xmath93 with @xmath94 non - empty .
because @xmath94 is non - empty , @xmath154 has length greater than 1 and lemma [ onelr1 ] applies to show that @xmath29 has a subsequence @xmath155 or @xmath156 for some @xmath157 .
to this sequence we can append the subsequence @xmath158 , take @xmath11 and @xmath5 in the roles of @xmath130 and @xmath131 and obtain a sequence of the required type .
@xmath101 and @xmath159 . here
@xmath160 is a subsequence of @xmath9 .
this is a case where the roles of @xmath132 and @xmath130 are both played by @xmath11 .
@xmath161 . here
@xmath162 is a subsequence of @xmath9 and @xmath11 again plays the roles of @xmath132 and @xmath130 .
[ twolr2 ] let @xmath43 be a permutation with exactly two left to right maxima and not ending in its largest element ( so of the type displayed in figure [ basispermextension ] ) .
suppose that @xmath9 contains a subsequence order isomorphic to a permutation in @xmath123 .
then @xmath26 contains a sequence order isomorphic to @xmath43 .
we use two principles that follow from lemma [ bdefinition2 ] regarding the transformation of @xmath9 into @xmath26 : 1 .
every subsequence of @xmath9 whose terms are not left to right maxima is transformed into the same subsequence , and 2 .
if @xmath163 is a subsequence of @xmath9 such that @xmath11 is the only left to right maxima of @xmath9 in the subsequence , then , in @xmath26 , the immediately preceding left to right maximum of @xmath9 lies between @xmath164 and @xmath165 .
let @xmath166 or @xmath167 be a subsequence of @xmath9 that is order isomorphic to a permutation of @xmath123 . here
@xmath168 is order isomorphic to @xmath43 .
indeed @xmath106 correspond to the black points in either diagram of figure [ basispermextension ] , @xmath134 correspond to the white points in the right - hand diagram , @xmath87 corresponds to @xmath17 , and @xmath125 corresponds to @xmath6 . by ( 1 ) above @xmath169 will be a subsequence of @xmath26 since no term in this subsequence of @xmath9 is a left to right maximum ( not even a left to right maximum of @xmath166 or @xmath167 ) .
also @xmath110 will precede @xmath169 in @xmath26 .
this follows from ( 1 ) if @xmath110 is not a left to right maximum of @xmath9 .
however , if @xmath110 is a left to right maximum , then @xmath110 precedes @xmath132 in @xmath9 ; but then @xmath132 is either a left to right maximum also , or there is another left to right maximum between @xmath110 and @xmath132 and we can appeal to ( 2 ) above . finally we show that , in @xmath26 , there is a term between @xmath87 and @xmath125 that , even if it is not @xmath5 itself , plays the role of @xmath5 in that it exceeds every term of @xmath125 . if @xmath5 is not a left to right maximum of @xmath9 then , by ( 1 ) , it will itself be positioned between @xmath87 and @xmath125 in @xmath26 .
so we suppose that @xmath5 is a left to right maximum of @xmath9 and let @xmath170 be the immediately preceding left to right maximum .
then , by ( 2 ) , @xmath170 will lie between @xmath87 and @xmath125 in @xmath26 so it is now sufficient to show that @xmath171 ( since @xmath172 ) .
indeed @xmath173 is impossible ; for either @xmath130 would precede @xmath170 and then @xmath170 would not be a left to right maximum , or @xmath170 would precede @xmath130 and then @xmath170 would not be the left to right maxima that immediately preceded @xmath5 .
hence @xmath26 contains @xmath43 .
now lemmas [ basicmethod ] , [ twolr1 ] and [ twolr2 ] prove [ twolrmain ] let @xmath43 be a permutation of length @xmath38 , with exactly two left to right maxima but not ending in its maximal symbol . then @xmath44 is a pattern class whose basis is the set of minimal permutations in @xmath123
. [ threelr1 ] if @xmath174 has exactly three left to right maxima , namely @xmath110 , @xmath5 and @xmath175 , and @xmath6 is non - empty then @xmath176 note that @xmath177 does not end with its maximum element and therefore we may use the same proof as in corollary [ aalphab ] .
there remains just one case : when @xmath178 has exactly three left to right maxima as in corollary [ threelr1 ] but when @xmath6 is empty . here
@xmath179 and @xmath180 ( written as @xmath181 and @xmath182 for typographical brevity ) .
this case is treated in the following proposition which is proved by the same approach followed in propositions [ onelrmain ] and [ twolrmain ] .
let @xmath43 be a permutation of the form @xmath183
. then @xmath184
this paper has characterised the principal pattern classes @xmath78 for which @xmath44 is a pattern class .
it has given a criterion in terms of left to right maxima of @xmath43 and we call such permutations `` good '' . for non - principal pattern classes
far less is known although the following easy result holds : of course this result proves not only that , when @xmath185 contains only good permutations , @xmath188 is a pattern class but it also allows its basis to be described ( as the set of minimal permutations in the union of bases of the pattern classes @xmath44 ) . pattern class research is often concerned with the enumeration question : how many permutations of length @xmath38 does a particular pattern class contain . for principal pattern classes @xmath78
no enumerations are known when @xmath190 and so it would be unrealistic to hope that many pattern classes of the form @xmath44 could be enumerated .
however a cruder question can be asked .
every pattern class @xmath191 has an upper growth rate @xmath192 defined as @xmath193a_n$ ] ( @xmath194 being the number of permutations in @xmath191 of length @xmath38 ) .
the set of permutations sortable with @xmath200 passes of bubble sort , namely @xmath201 , is a pattern class whose basis is the set of @xmath202 permutations of length @xmath203 whose final term is @xmath112 .
the proof of this result has a slight bearing on problem [ necsuff ] .
the set @xmath207 contains @xmath208 which is not a good permutation .
nevertheless @xmath209 is a pattern class .
hence the necessary and sufficient condition asked for in problem [ necsuff ] is more subtle than the condition that @xmath185 consists entirely of good permutations .
finally , we note that the pattern classes featuring in these propositions have been enumerated .
our calculations show that the number of permutations of length @xmath38 in @xmath198 is @xmath210 and it is well - known that there are @xmath211 permutations of length @xmath38 in @xmath212 .
99 s. dulucq , s. gire , o. guibert : a combinatorial proof of j. west s conjecture , discrete mathematics 187 ( 1998 ) , 7196 .
m. bousquet - mlou : multi - statistic enumeration of two - stack sortable permutations , electronic journal of combinatorics 5 ( 1998 ) , paper r21 . j. west : sorting twice through a stack , theoretical computer science 117 ( 1993 ) , 303313 .
d. zeilberger : a proof of julian west s conjecture that the number of two - stack - sortable permutations of length @xmath38 is @xmath213 , discrete mathematics 102 ( 1992 ) , 8593 . | let @xmath0 be the operation of re - ordering a sequence by one pass of bubble sort .
we completely answer the question of when the inverse image of a principal pattern class under @xmath0 is a pattern class . |
operator splitting is a widely used time discretization method for the numerical solution of complicated equations .
the importance and main applications of these procedures is described , for example , in the monographs by farag and havasi @xcite , holden et al .
@xcite and lubich @xcite .
the present paper investigates a special operator splitting for a class of nonautonomous delay differential equations .
this method , which can be applied to equations with distributed delays very effectively , was first investigated in csoms and nickel @xcite and in btkai , csoms and nickel @xcite in the autonomous case .
recall from the introduction in bellen and zennaro @xcite that delay equations with distributed delay , especially those where the delay term is not separated from zero , are particularly difficult to solve numerically . nevertheless , as we shall see
, splitting methods work quite well even in the latter case . to motivate this approach ,
let us consider the following equation .
@xmath0 , \end{aligned } \right .
\nonumber\ ] ] where @xmath1 , @xmath2)$ ] , and @xmath3 for all @xmath4 $ ] . in this case
the delay operator @xmath5 ( cf .
equation below ) is defined by @xmath6 for all @xmath7)$ ] . choosing a time step @xmath8 $ ] ,
first we start with @xmath9 and @xmath10
. then we set @xmath11 and @xmath12 , \\
f_0(h+\sigma),\quad \sigma\in [ -1,-h).\end{cases}\ ] ] in the next step we repeat this procedure and replace @xmath13 with @xmath14 , @xmath15 with @xmath16 , and @xmath17 with @xmath18 .
hence , we obtain an iteration process where in the @xmath19 step we have @xmath20 , \\
f_{k-1}(h+\sigma),\quad \sigma\in [ -1,-h).\end{cases } \end{aligned } \right.\ ] ] the aim of the present paper is to prove the convergence of this procedure also for more general equations .
we do this by introducing an abstract setup allowing us a general convergence result of the procedure .
further , for differentiable initial function ( i.e. , classical solutions ) we also obtain estimates on the order of convergence . in the following ,
we summarize some basic facts on product formulae for abstract evolution equations . then in section 2 we rewrite the nonautonomous delay equation as an abstract evolution equation and prove the convergence of a general product formula .
section 3 is devoted to the investigation of the order of convergence , and in section 4 we present numerical examples demonstrating the power of this approach .
first , let us recall some general facts about splitting of nonautonomous equations .
consider an evolution equation of the form @xmath21 we always suppose that this equation is well - posed , i.e. , there is an evolution family @xmath22 ( also called semi - dynamical system ) solving it . by an _ evolution family
_ we mean a strongly continuous two - parameter family of bounded linear operators @xmath22 with the algebraic property @xmath23 and @xmath24 for all @xmath25 . for well - posedness of nonautonomous evolution equations we refer to the surveys in nagel and nickel @xcite ,
pazy ( * ? ? ?
* chapter 5 ) or schnaubelt in ( * ? ? ? * chapter vi.9 ) .
now , a splitting formula , as below , is especially useful , if we are able to solve effectively the _ autonomous _ cauchy problems @xmath26 with appropriate initial conditions for every fixed @xmath27 .
this is usually the case , if the operators @xmath28 and @xmath29 are partial differential operators with time dependent coefficients , or time dependent multiplication operators , and this is particulary so for delay equations considered in this paper .
the following general convergence result can be proved , see btkai et al .
* theorem 4.2 ) .
[ th : firstproduct ] suppose the following : a. the nonautonomous cauchy problem corresponding to the operators @xmath30 is well - posed .
we denote the evolution family solving by @xmath22 .
b. the operators @xmath31 and @xmath32 are generators of @xmath33-semigroups of type @xmath34 ( @xmath35 and @xmath36 ) , and @xmath37 and @xmath38 c. the maps @xmath39 are continuous for all @xmath40 and @xmath41 .
then one has the convergence @xmath42 for all @xmath41 , locally uniformly in @xmath43 with @xmath44 .
consider the * abstract delay equation * in the following form : @xmath45;x ) \end{aligned } \right .
\label{delay}\ ] ] on the banach space @xmath46 , where @xmath31 generates a strongly continuous contraction semigroup on @xmath46 and @xmath47;x ) \to x$ ] is a bounded and linear operator depending continuously on the parameter @xmath48 .
the * history function * @xmath49 is defined by @xmath50 for @xmath4 $ ] .
note that point delays are excluded from this context , but _ distributed delays _
, even those that live up to @xmath51 , are contained in this setting . in order to rewrite as an abstract cauchy problem ,
we take the product space @xmath52;x)$ ] equipped with @xmath53-sum norm , and the new unknown function as @xmath54 then can be written as an abstract cauchy problem on the space @xmath55 in the following way : @xmath56 where the operator @xmath57 is given by the matrix @xmath58 on the domain @xmath59;x ) : \,f(0)=x \right\}. \nonumber\ ] ] as in btkai and piazzera ( * ? ? ?
* corollary 3.5 , proposition 3.9 ) one can show that the delay equation and the abstract cauchy problem are equivalent , i.e. , they have the same solutions .
more precisely , the first coordinate of the solution of always solves , i.e. , @xmath60 where @xmath61 is the projection to the first coordinate in @xmath55 . due to this equivalence ,
the delay equation is well - posed if and only if the operator @xmath57 generates an evolution family on the space @xmath55 .
since the delay operators @xmath5 are bounded , the delay equation is well - posed , which follows form a much more general well - posedness result by hadd , rhandi and schnaubelt ( * ? ? ?
* proposition 3.5 ) .
that is , there is an _ evolution family _
@xmath62 such that for fixed @xmath17 and @xmath63 the function @xmath64 is a solution of for @xmath65 .
in particular , for @xmath66 we have @xmath67 $ , $ t+r\geq s$}.\end{aligned}\ ] ] furthermore , we have the next relation @xmath68 now we make the main assumptions implying the convergence of the splitting procedure . in the autonomous case ,
i.e. , when @xmath69 and @xmath70 , the following was investigated in the papers by csoms and nickel @xcite and btkai , csoms , and nickel @xcite .
[ ass:2 ] a. the operators @xmath71 generate the strongly continuous contraction semigroups @xmath72 on @xmath46 for all @xmath73 .
b. @xmath74 for all @xmath75 and the function @xmath76 is continuous for all @xmath41 .
c. the delay operators @xmath77;x ) \to x$ ] are bounded for all @xmath75 . d. the function @xmath78 is bounded and continuous for every @xmath79;x)$ ] .
[ ass1 ] let us now describe in detail the approximation procedure we will apply . we split the operator in as @xmath80 where the sub - operators have the form @xmath81 since @xmath28 is a generator and @xmath82 is bounded , the operators @xmath83 and @xmath84 generate the strongly continuous semigroups @xmath85 and @xmath86 , respectively .
it is shown in btkai and piazzera ( * ? ? ?
* theorem 3.25 ) that @xmath87 is given by @xmath88 where @xmath89 is the nilpotent left shift semigroup defined by @xmath90 \end{cases } \nonumber\ ] ] for all @xmath91;x)$ ] , and @xmath92 is @xmath93 , \\ 0 , & \quad\mbox{if}\quad \sigma\in[-1,-t ) \end{cases } \nonumber\ ] ] for all @xmath94 .
since @xmath82 is a bounded operator , @xmath84 is also bounded on @xmath55 .
therefore , the semigroup @xmath95 generated by @xmath84 takes the form @xmath96 where @xmath97 , @xmath98 , and @xmath99 denote the identity operators on @xmath46 , @xmath100;x)$ ] , and @xmath55 , respectively .
we then have the following general convergence result explaining the convergence of the procedure described in .
[ thm : general ] under assumption [ ass1 ] the solution of the abstract delay equation is given by the formula @xmath101 the convergence is a direct consequence of theorem [ th : firstproduct ] applied to the generators @xmath83 and @xmath84 .
as mentioned above , by hadd , rhandi and schnaubelt @xcite the cauchy problem associated to the operator @xmath102 is well - posed .
applying theorem [ th : firstproduct ] , we only have to check the stability assumption . indeed ,
assumption [ ass:2].a ) and b ) and an argument using power series imply the continuity of @xmath103 for @xmath104 in the connected component of @xmath53 , hence condition c ) on theorem [ th : firstproduct ] is fulfilled .
now , to see stability one can use the same arguments as the ones appearing in the proof of csoms and nickel ( * ? ? ?
* theorem 4.2 ) , we get that @xmath105 hence @xmath106 from this the stability can be obtained : @xmath107 with @xmath108 .
we now investigate the order of convergence of the splitting method from the previous section . to this end
, we have to make further assumptions on the operators involved , and of course , some regularity on the initial data need to be assumed , too . [ ass : convrate ] a.
the operator @xmath71 is bounded and generates the strongly continuous contraction semigroup @xmath72 on @xmath46 . b. the delay operators @xmath77;x ) \to x$ ] are bounded for all @xmath75 . c. the function @xmath109 is bounded and locally lipschitz continuous , i.e. , for all @xmath110 there is @xmath111 such that @xmath112 for all @xmath113 .
d. the function @xmath78 is bounded and locally lipschitz continuous , i.e. , for all @xmath110 there is @xmath111 such that @xmath114 for all @xmath113 .
these assumptions enable us to show the first order of convergence for classical solutions .
[ thm : approxt2 ] let @xmath115 be fixed .
then there is a constant @xmath116 such that @xmath117 holds for all @xmath118 $ ] , @xmath119 $ ] and @xmath120;x)$ ] , @xmath121 .
recall from the previous section that for @xmath122 we have @xmath123 from this it follows that @xmath124\to x$ ] is lipschitz continuous with constant @xmath125 with @xmath126 dependent only on @xmath127 and @xmath128 .
let us calculate the product @xmath129 and compare the first component here with @xmath130 .
we can write @xmath131 and by writing out the series expansion of @xmath132 we obtain @xmath133 where @xmath134 denotes a term bounded in norm by @xmath135 with a constant @xmath136 that depends only on the bounds of @xmath127 and @xmath128 . we now can write @xmath137 where @xmath138 is the lipschitz constant of @xmath139 on @xmath140 $ ] . a very similar reasoning works for the other two terms in : @xmath141}\|_\infty\bigr\}\tfrac{h^2}2 .
\intertext{since $ f\in \w^{1,1}([-1,0];x)$ , we can continue the estimation : } & \bigl\|t\phi(s ) f-\int_s^{s+h } \phi(r)u^{(s)}_r\dd r\bigr\|\leq \tfrac{l'}{2}\|f\|_1h^2+\|\phi\|_\infty\int_s^{s+h}\|f'\|_1(r - s ) \dd r\\ & \qquad\qquad\qquad+\|\phi\|_\infty\max\bigl\{\|f\|_\infty,\|u^{(s)}|_{[s , s+1]}\|_\infty\bigr\}\tfrac{h^2}{2}\\ & \quad\leq c ( \|x\|+\|f\|_1+\|f'\|_1)h^2.\end{aligned}\ ] ] by summing up we obtain the estimate : @xmath142 hence the assertion for the first coordinate in is proved .
let us now turn our attention to the second coordinate of . for @xmath143 $ ] we have the following :
@xmath144 we estimate the @xmath145-norm of @xmath146 by using the pointwise estimate for the integrand proved above for the first coordinate , we obtain that @xmath147 the proof is hence complete .
we now can prove the first order convergence of the sequential splitting .
passing from local error estimates to convergence is done by the standard trick of telescopic summation . for every @xmath115
there is constant @xmath116 such that for all @xmath120;x)$ ] and @xmath121 the inequality @xmath148 holds for all @xmath119 $ ] , @xmath149 $ ] and for all @xmath150 , where @xmath151 .
take @xmath152 so large that @xmath153 holds .
fix @xmath154 , @xmath155 , and @xmath156 as in the assertion . by telescopic summation
we obtain @xmath157\times\\ \notag&\qquad\qquad\times \bigl(\prod_{j=0}^{k-1}\mathcal{w}(s+(j+1)h , s+jh)\bigr)\tvect{x}{f}\biggr)=\\ \notag&\qquad=\sum_{k=0}^{n-1}\biggl(\bigl(\prod_{j = k+1}^{n-1 } \calt^{(s+ jh)}(h)\cals^{(s+ jh)}(h)\bigr)\times\\
\notag&\qquad\qquad\times\bigl[\calt^{(s+ kh)}(h)\cals^{(s+ kh)}(h)-\mathcal{w}(s+(k+1)h , s+kh)\bigr]\times\\ \notag&\qquad\qquad\times \mathcal{w}(s+kh , s)\tvect{x}{f}\biggr).\end{aligned}\ ] ] for @xmath158 we have @xmath159
. therefore we can conclude from theorem [ thm : approxt2 ] that @xmath160 holds for all @xmath161 .
from this and from it follows @xmath162 hence the assertion .
in this section we present some numerical experiments obtained by the scheme described in theorem [ thm : general ] and in . the program code is a modification of the code described in csoms and nickel @xcite . in order to check the convergence of the numerical scheme in the nonautonomous case , and compare the solutions in the autonomous and nonautonomous cases
, we will investigate the following examples .
let @xmath163 , @xmath164 and consider @xmath165;\rr ) , \end{aligned } \right.\ ] ] for some @xmath2)$ ] . in this case the delay operator @xmath5 is defined by @xmath166 for all @xmath7;\rr)$ ] .
let us choose the initial values as @xmath167 and @xmath168 for @xmath4 $ ] , and @xmath169 . as a particular example we choose the following functions @xmath170 : a. @xmath171 in the autonomous case , and b. @xmath172 in the nonautonomous case , for @xmath173 and @xmath4 $ ] .
[ exm:1 ] though point delays do not fit into the framework of the theoretical part of this paper , it is worth to investigate how this splitting method performs for them .
let us consider @xmath163 , @xmath164 and @xmath174;\rr ) . \end{aligned } \right .
\nonumber\ ] ] the delay operator in this case is @xmath175 for all @xmath176;\rr)$ ] . let us choose the initial values again as @xmath167 and @xmath168 for @xmath177 , and @xmath169 .
as above , we consider again the functions
@xmath170 : a. @xmath171 in the autonomous case , and b. @xmath172 in the nonautonomous case , for @xmath173 and @xmath4 $ ] .
[ exm:2 ] in order to examine the convergence of the numerical scheme described in theorem [ thm : general ] , we plot the values of the split solution @xmath178 obtained by the formula @xmath179 for different values of time steps @xmath156 .
the results for the nonautonomous case are shown on figure [ fig : conv ] .
one can see that the split solutions converges to the exact solution if @xmath156 decreases .
the convergence of this numerical scheme in the autonomous case has been already investigated in csoms , nickel ( * ? ? ?
* section 5.3 ) . [ cols="^,^ " , ]
a. b. was supported by the alexander von humboldt - stiftung and by the otka grant nr . k81403 .
the european union and the european social fund have provided financial support to the project under the grant agreement no .
tmop-4.2.1/b-09/1/kmr-2010 - 0003 . during the preparation of the paper b.
f. was supported by the jnos bolyai research scholarship of the hungarian academy of sciences . the financial support of the `` stiftung aktion sterreich - ungarn '' is gratefully acknowledged .
engel and r. nagel , _ one - parameter semigroups for linear evolution equations _ , graduate texts in mathematics , vol .
194 , springer - verlag , new york , 2000 , with contributions by s. brendle , m. campiti , t. hahn , g. metafune , g. nickel , d. pallara , c. perazzoli , a. rhandi , s. romanelli and r. schnaubelt .
h. holden , k. h. karlsen , k .- a .
lie , and n. h. risebro , _ splitting methods for partial differential equations with rough solutions _ , ems series of lectures in mathematics , european mathematical society ( ems ) , zrich , 2010 , analysis and matlab programs .
r. nagel and g. nickel , _ well - posedness for nonautonomous abstract cauchy problems _ , evolution equations , semigroups and functional analysis ( milano , 2000 ) , progr .
nonlinear differential equations appl .
50 , birkhuser , basel , 2002 , pp . | we provide a general product formula for the solution of nonautonomous abstract delay equations .
after having shown the convergence we obtain estimates on the order of convergence for differentiable history functions .
finally , the theoretical results are demonstrated by some typical numerical examples . |
the revolution in the understanding of the present universe has become possible due to the research in observational cosmology in the past two decade .
recent observational data from plank collaboration @xcite , baryon oscillation spectroscopic survey ( boss ) @xcite and atacama cosmology telescope polarimeter ( actpol ) collaboration @xcite provides relevant experimental evidence about the acceleration of our universe . in the last century
, modern cosmology reached a new vision to establish considerable advancements in the account of the current accelerated expanding universe .
the two crucial observational groups including supernovae cosmology project and the high - redshift supernovae search team have provided the main evidence for the cosmic acceleration of the universe @xcite .
the other cosmic observations like cosmic microwave background ( cmb ) fluctuations @xcite , large - scale structure ( lss ) @xcite , cosmic microwave radiation ( cmbr ) @xcite indicate that the present universe is undergoing an accelerated expansion .
it is also believed that the universe changed with time from early deceleration phase to late time acceleration phase @xcite .
now , a team of scientists led by professor subir sarkar @xcite has raised the question on this standard cosmological concept about the acceleration , which is also supported by shariff et al .
+ the two promising approaches confirmed by the cosmological research community for discussing the cosmic expansion of the universe are : the first one is the introduction of the most exotic entity of mysterious universe dubbed as dark energy , which has positive energy density and negative pressure .
recently , from plank cosmological results and wilkson microwave anisotropic probe ( wamp ) 9 years results @xcite , it may be concluded that the universe embodied with 68.5% dark energy , 26.5% of dark matter and 5% of baryonic matter .
dark energy can be represented either in terms of cosmological constant or by the help of equation of state parameter ( eos ) @xmath5 , where @xmath6 is the pressure and @xmath7 is the energy density .
+ the second approach to picturize the evolution of the universe is a modified version of einstein s field equations of general relativity , which can be established through the einstein - hilbert action principle . in this process , the matter lagrangian is replaced by an arbitrary function .
thereafter , these modified theories become most attractive aspirant to observe the accelerated expansion of the universe as well as the effective causes related to dark energy .
+ in addition , it has been postulated that the standard einstein - hilbert action is modified by an arbitrary function @xmath8 , where @xmath1 is ricci scalar curvature .
the @xmath8 gravity become an adequate theory to provide the gravitational alternative for dark energy and about the early inflation plus late - time cosmic acceleration of the universe @xcite . in 2007 ,
the @xmath8 gravity theory is restructured by merging the matter lagrangian density @xmath9 with initial arbitrary function of the ricci scalar @xmath1 @xcite .
the unification of dark energy and early time inflation with late time acceleration from @xmath8 theory to all lorentz non - invariant theories is discussed by nojiri and odintsov @xcite . through continuation of this work of coupling , in 2011 , harko et al .
@xcite proposed a new modified theory named as @xmath0 theory , where the gravitational part of the action still depends on the ricci scalar @xmath1 like @xmath8 theories and also a function of trace @xmath2 .
it is suggested that due to the matter - energy coupling , the leading model of this theory depends on source term representing the variation of energy - momentum tensor .
indefinitely many modified gravitational theories such as @xmath10 gravity , @xmath11 gravity , and @xmath12 gravity etc . were developed to achieve the accelerated expansion of the universe .
myrzakulov et al .
@xcite have investigated the inflation in @xmath13 theories of gravity where the scalar field is coupled with gravity .
sebastiani and myrzakulov @xcite have briefly reviewed various @xmath8 gravity models for inflation , in particular , starobinsky - like inflation .
after that , @xmath0 gravity become most prominent theory for investigating the fate of the late time accelerating expansion of the universe .
a phase transition also occurred from matter - dominated era to an accelerated phase during the reconstruction of @xmath0 gravity theory @xcite .
in the context of common perfect fluid matter , an axially symmetric cosmological model was constructed in the framework of @xmath0 gravity @xcite . in @xmath0 gravity theory , many cosmological models can be constructed changing choices in matter source .
sahoo and mishra @xcite constructed kaluza - klein dark energy model by considering the wet dark fluid matter source .
recently , moraes et al .
@xcite derived the starobinksy model in @xmath0 gravity .
+ by considering the metric dependent lagrangian density @xmath9 , the respective field equation for @xmath0 gravity is formulated from the hilbert - einstein variational principle in the following manner : @xmath14where , @xmath15 is the usual matter lagrangian density of matter source , @xmath0 is an arbitrary function of ricci scalar @xmath1 and the trace @xmath2 of the energy - momentum tensor @xmath16 of the matter source , and @xmath17 is the determinant of the metric tensor @xmath18 .
the energy - momentum tensor @xmath16 from lagrangian matter is defined in the form @xmath19and its trace is @xmath20.here , we have assumed that the matter lagrangian @xmath15 depends only on the metric tensor component @xmath18 rather than its derivatives .
hence , we obtain @xmath21by varying the action @xmath22 in eq .
( 1 ) with respect to @xmath18 , the @xmath0 gravity field equations are obtained as @xmath23where , @xmath24 here , @xmath25 , @xmath26 , @xmath27 where @xmath28 is the covariant derivative.contracting eq .
we get @xmath29where @xmath30.from eqs ( 4 ) and ( 6 ) , the @xmath0 gravity field equations takes the form @xmath31it is worth mentioning here that the physical nature of the matter field through @xmath32 is used to form the field equations of @xmath0 gravity . to construct different kinds of cosmological models according to the choice of matter source , harko et al .
@xcite constructed three types of @xmath0 gravity as presented below @xmath33the individual field equations for each frames of @xmath0 gravity is given as + case - i : @xmath3 @xmath34
case - ii : @xmath4 @xmath35 most of the researchers constructed cosmological models in the presence of perfect fluid matter to analyse the accelerated expansion of the universe .
recent observations provide the evidence for the accelerated expansion of the universe due to the presence of an unknown form of energy ( dark energy ) having negative pressure . on that account
, we need to construct a cosmological model for expanding universe without considering dark energy and dark matter , while choosing most reliable matter component .
it is believed that the cosmic viscosity acts as the dark energy candidate which may play an important role in causing the accelerated expansion of the universe by consuming negative effective pressure @xcite . at present , there is great interest in formulating the cosmological model with dissipative fluid matter rather than with dust ( a pressure - less distribution ) or with a perfect fluid matter , which gives more realistic model instead of others and most effective to pay attention to the dynamical background of homogeneous and isotropic universe .
it is commonly accepted that , in the early phase of the universe , the matter behaved like a viscous fluid during the neutrino decoupling in radiation era @xcite .
the nature of singularity occurred for perfect fluid can be modified through the dissipative mechanism of viscous fluid .
this viscous fluid cosmological model helps to explain the matter distribution on the large entropy per baryon in the present universe .
also , the phase transition and string creation are involved with viscous effects as per the grand unified theories ( gut ) .
it has been found that the mixture of minimally coupled self - interacting scalar field can successfully derive an accelerated expansion of the universe , while the same mixture with perfect fluid unable to do it @xcite .
hence , many cosmological models with viscous fluid in the early universe have been widely discussed in the literature @xcite .
again , observations have been conducted to obtain the homogeneity and isotropic properties of the universe .
it is believed that at the end of the inflationary era , the geometry of the universe was homogeneous and isotropic @xcite , where the flrw models played an important role is representing both spatially homogeneous and isotropic universe . but the theoretical argument and the anomalies found in cmb provide the evidence for the existence of an anisotropic phase , which is later called isotropic one . after the announcement of plank probe results @xcite , it is believed that the early universe may not have been exactly uniform .
thus , the existence of inhomogeneous and anisotropic properties of the universe has gained popularity when it comes to constructing cosmological models under the supervision of anisotropic background .
therefore , bianchi type models are very relevant for describing the early universe with the anisotropic background . due to some analytical difficulties in studying the inhomogeneous models , many researchers considered the bianchi type models for investigating the cosmic evolution of the early universe as they are homogeneous and anisotropic .
there exist nine types ( @xmath36 ) of bianchi space - times in literature . here , we consider bianchi type - i space - time , as it is the simplest spatially homogeneous and anisotropic .
it is also known as the immediate generalization of the flrw flat metric with different scale factors in each spatial direction . in some special cases ,
the bianchi type - i models include _ kasner metric _ , which helps to govern the dynamics near the singularity .
the bianchi type - i cosmological models are more compatible with the simplest mathematical form which attracts various researcher to study different aspects .
the nature of bianchi type - i cosmological model has been studied in the context of a viscous fluid .
it has been observed that viscosity can cause the qualitative behaviour of solutions near the singularity without removing the total initial big bang singularity @xcite .
recently , the bulk viscous matter content has been discussed along with cosmological constant in the framework of bianchi type - i space time @xcite .
later , the spatially homogeneous anisotropic bianchi type - i model with bulk viscous fluid matter was studied with the assumption of constant deceleration parameter in the framework of @xmath0 gravity theory through the explanation of the energy conditions @xcite.this work is motivated by the aforesaid literature .
the study is presented in different sections .
section - i deals with the basic formalism of @xmath0 gravity field equations from einstein - hilbert action along with some essential literature review . in section - ii
, we derive the exact solution of both the cases of @xmath0 gravity ( @xmath3 and @xmath4 ) for the spatially homogeneous anisotropic bianchi type - i space - time with the help of time varying deceleration parameter . the energy density ,
bulk viscous pressure , bulk viscous coefficient , the trace of matter , ricci scalar , the energy conditions are derived and their graphical representations are also discussed in this section . in section - iii ,
physical parameters of both the models are presented .
the detail discussion of the figures are covered in section - iv .
the last section contains the conclusion of the covered study .
we consider the spatially homogeneous lrs bianchi type - i metric as @xmath37 where @xmath38 are functions of cosmic time @xmath39 only .
+ the energy momentum tensor for bulk viscus fluid is considered in the form @xmath40 where @xmath41 is the four velocity vector in co - moving coordinate system satisfying @xmath42 @xmath43 is the bulk viscous pressure which satisfies the linear equation of state @xmath44 , @xmath45 is the bulk viscous coefficient , @xmath46 is hubble s parameter , @xmath6 is pressure and @xmath7 is the energy density . + the trace of energy momentum tensor is given as @xmath47 the field equation ( 9 ) with @xmath48 , where @xmath49 is an arbitrary constant for the metric ( 11 ) is obtained as @xmath50 dot represent derivatives with respect to time @xmath51 + the deceleration parameter ( dp ) is defined as @xmath52 where , @xmath53 is the average scale factor .
+ we have three equations ( 15 - 17 ) involving four parameters as @xmath54 . in order to solve these equations
, we assume the time varying deceleration parameter as @xmath55 where , @xmath56 is a constant .
+ the hubble s parameter is defined as @xmath57 , and from above equation we obtain @xmath58 where , the integrating constant is assumed as unity . integrating ( 20 )
we have found @xmath59 setting @xmath60 , where @xmath61 is the redshift , leads to relation @xmath62 the corresponding @xmath63 is obtained as @xmath64 the deceleration parameter describes the evolution of the universe .
the cosmological models of the evolving universe transits from early decelerating phase ( @xmath65 ) to current accelerating phase ( @xmath66 ) . whereas , the models can be classified on the basis of the time dependence of dp .
recent observations like sne ia @xcite and cmb anisotropy @xcite confirmed that the present universe is undergoing an accelerated phase of expansion and the value lies in between @xmath67 .
fig-1 depicts the behavior of deceleration parameter with respect to redshift , in which the value of @xmath68 lies in specified range of accelerating phase .
the transition phase of deceleration parameter from deceleration to acceleration occurs at some transition redshift @xmath69 which is completely decided by the parameter @xmath70 . in this work ,
we have considered three representative values of @xmath70 i.e. 1.4772 , 1.5 and 1.55 corresponding to @xmath71 , 0.5874 and 0.4706 respectively .
the values of transition redshift @xmath69 for our model are agreeing with the observational data @xcite .
the transition from deceleration to acceleration phase in @xmath0 gravity with polynomial function of @xmath2 is discussed by moraes et al .
our first case clearly shows that the model is completely under accelerated phase which is conformity with observational data . vs. redshift with different @xmath70.,width=283 ]
the volume is defined as @xmath72 . using ( 21 ) in this relation , the values of the metric potentials @xmath38
are obtained as @xmath73 consequently metric ( 11 ) takes the form @xmath74 solving the field equations ( 15 - 17 ) , the values of @xmath7 and @xmath75 are obtained as @xmath76\ ] ] @xmath77\end{gathered}\ ] ] we can observe that the energy density remains positive throughout the evolution of the universe and is a decreasing function of cosmic time @xmath39 .
it starts with a positive value and approaches to zero as @xmath78 .
the bulk viscous pressure @xmath75 is an increasing function of time , which begins from a large negative value and tends to zero at present epoch .
as per the observation , the negative pressure is due to de in the context of accelerated expansion of the universe .
hence , the behavior of bulk viscous pressure in our model is agreed with this observation .
+ the coefficient of bulk viscosity @xmath45 and the pressure are obtained as @xmath79\end{gathered}\ ] ] @xmath80\ ] ] vs. time with @xmath81 and different @xmath70.,width=245 ] vs. time with @xmath81 and different @xmath70.,width=245 ] vs. time with @xmath81 and different @xmath70.,width=245 ] figure 4 , indicates that the bulk viscous coefficient @xmath45 is positive through out the universe and becomes finite as @xmath78 for our model .
+ the energy conditions are some alternative conditions for matter content of the theory . in general relativity ( gr ) , the role of these energy conditions is to prove the theorems about the existence of space - time singularity and black holes @xcite .
the energy conditions are used in many approaches to understand the evolution of the universe . here , we discussed some of the most popular energy conditions for this model .
the weak energy conditions ( wec ) , dominant energy conditions ( dec ) and strong energy conditions ( sec ) are given as @xmath82 these are used in various aspects through their own importance .
for example , the dec gives an idea about the stability of matter source and imposes the dark energy along with equation of state parameter @xmath83 for lower bound @xmath84 , which may cause big rip @xcite .
sec violation is a typical trait of a positive cosmological constant @xmath85 @xcite .
finally , wec shows that the matter - energy is always non - negative . in @xmath0 gravity ,
the matter support of wormholes is examined by the behavior of energy conditions and the nature of solution for frw model with perfect fluid matter is studied through energy conditions @xcite . according to the predefined literature
, one can consider the energy conditions to be useful to analyze the behavior of cosmological solutions throughout the universe .
therefore , we dealt with some well - known energy conditions like wec , dec and sec to observe our solutions in both of the cases . figures(5 - 7 ) show the behavior of wec , dec , and sec with the proper choice of constants respectively .
+ and @xmath70 with @xmath86.,width=245 ] and @xmath70 with @xmath86.,width=245 ] and @xmath70 with @xmath86.,width=245 ] from figures 5 - 7 , it is observed that all the energy conditions are satisfied for this model .
+ the values of ricci scalar @xmath1 and the trace of matter source @xmath2 are obtained as @xmath87 = \biggl(6\beta-\frac{27}{2}\biggr)e^{2\beta t}(e^{\beta t}-1)^{-2}-6\beta e^{\beta t}(e^{\beta t}-1)^{-1}\ ] ] @xmath88\end{gathered}\ ] ] using the above equations , the function @xmath0 is obtained as @xmath89 versus @xmath39 and @xmath70 with @xmath90.,width=283 ] fig .
8 shows the behaviour of the function @xmath0 for this model .
in this case we assumed @xmath91 and @xmath92 , where @xmath93 is an arbitrary constant .
the corresponding field equations in the general form is @xmath94 where @xmath95 .
the set of field equations for the metric ( 11 ) are @xmath96 the above set of field equations ( 38 - 40 ) admits the same solutions ( 24 ) and ( 25 ) as obtained in the previous case . using the metric potentials from ( 24 ) and ( 25 ) , the values of energy density @xmath7 and bulk viscous pressure @xmath75 are @xmath97\ ] ] @xmath98\end{gathered}\ ] ] fig .
9 shows that the energy density @xmath7 is decreasing function of time and remains always positive .
it converges to zero as @xmath78 .
10 depicts the variation of @xmath75 versus cosmic time @xmath39 .
it is observed from the figure that bulk viscous pressure is increasing function of time from a large negative value and is approaching to zero at present time .
+ for this case , the values of bulk viscosity coefficient @xmath45 and effective pressure @xmath6 are @xmath99\end{gathered}\ ] ] @xmath100\ ] ] vs. time with @xmath101 and different @xmath70,width=245 ] vs. time with @xmath101 and different @xmath70,width=245 ] vs. time with @xmath101 and different @xmath70,width=245 ] as suggested in figure 11 , the bulk viscosity coefficient is constant throughout the universe as required .
+ the weak energy conditions ( wec ) , dominant energy conditions ( dec ) and strong energy conditions ( sec ) for this model are plotted below .
+ and @xmath70 with @xmath101.,width=245 ] and @xmath70 with @xmath101.,width=245 ] and @xmath70 with @xmath101.,width=245 ] figures 12 - 14 , show that the energy conditions are completely agreed with gr .
+ we can obtain the trace of matter @xmath2 for this model as @xmath102\end{gathered}\ ] ] the relation @xmath0 for the above case is obtained in the form @xmath103 versus @xmath39 and @xmath70 with @xmath101.,width=283 ] fig .
15 shows the behaviour of @xmath0 for @xmath4 model .
the rate of expansion of the universe with respect to time is defined by hubble s parameter as well as deceleration parameter . for detail
kinematical descriptions of the cosmological expansions can be obtained by taking in account of some extended set of parameters having higher order time derivatives of the scale factor .
+ the spatial volume turns out to be @xmath104 the above equation indicates that in both the models the spatial volume is zero at initial time @xmath105 .
it shows that the evolution of our universe starts with big bang scenario .
it is further noted that from ( 21 ) the average scale factor becomes zero at the initial epoch .
hence , both models have a point type singularity @xcite . the spatial volume increases with time . + the hubble s parameter @xmath46
, expansion scalar @xmath106 and shear scalar @xmath107 become @xmath108 @xmath109 @xmath110 from the above equations , we can observe that the hubble factor , scalar expansion and shear scalar diverge at @xmath105 and they become finite as @xmath78 .
it is noted here that the isotropic condition @xmath111 becomes constant ( from early to late time ) , which shows that the model does not approach isotropy throughout the evolution of the universe .
the anisotropy parameter @xmath112 the anisotropic parameter becomes constant for our models . from the above mentioned equation
it can be observed that our models are expanding and accelerating universe which starts at a big bang singularity . +
* jerk parameter : * + the jerk parameter is considered as one of the important quantity for describing the dynamics of the universe .
the models close to @xmath85cdm can be described by the cosmic jerk parameter @xmath113 @xcite . for flat @xmath85cdm model
the value of jerk is @xmath114 @xcite .
jerk parameter is a dimensionless third derivative of scale factor @xmath53 with respect to cosmic time @xmath39 and is defined as @xmath115 the above expression can be written in terms of deceleration parameter as @xmath116 thus , the jerk parameter for our models is @xmath117 with different @xmath70.,width=283 ] from the fig .
16 , it is clear that our value does not overlap with the value @xmath118 obtained from combination of three kinematical data sets : the gold sample data of type ia supernovae @xcite , the snia data obtained from the snls project @xcite , and the x - ray galaxy cluster distance measurements @xcite .
we have plotted the jerk parameter for different values of @xmath70 in fig .
. one can observe that the jerk parameter remains positive through out the universe and is equal to the @xmath85cdm model at @xmath119 for the considered values of @xmath70 .
it is interesting to note that our model is close to @xmath85cdm model for the following set of values as presented in table - i . + * @xmath120 parameter : * + the state - finder pair @xmath121 is defined as @xcite @xmath122 the state - finder pair is a geometrical diagnostic parameter , which is constructed from a space - time metric directly , and it is more universal compared to physical variables , which depend on the properties of physical fields describing de , since physical variables are model dependent . for the flat @xmath85cdm model
the state - finder pair obtained as @xmath123 @xcite .
the values of the state - finder parameter for our model are @xmath124 @xmath125\ ] ] from the expressions of @xmath126 and @xmath127 parameters , we found that @xmath123 only when @xmath128 .
the variation of @xmath70 and @xmath39 for @xmath123 is presented in table - i .
for the set of values of @xmath129 our models represents @xmath85cdm models , which are presented in table - i . the @xmath120 trajectory of our models is presented in the fig .
.variation of @xmath70 and @xmath39 for @xmath121 [ cols="^,^,^,^",options="header " , ] from the above table - i , it is observed that at initial epoch @xmath105 the parameter @xmath126 becomes unity and @xmath127 becomes finite and diverges for @xmath130 . vs. @xmath127.,width=283 ]
it is well known that the standard @xmath85cdm ( cosmological constant @xmath85+cold dark matter ) model fits well to the currently available observational data .
the most recent data from planck+bao+jla+@xmath131 yields the matter density parameter @xmath132 in the @xmath85cdm model @xcite . using this value , we find that the @xmath85cdm universe transits from deceleration to acceleration at the redshift @xmath133 . in our models
, it corresponds to @xmath134 as may be seen in figure 1 , where we show the variation of deceleration parameter vs redshift for @xmath134 .
thus the models present here have a transition from deceleration to acceleration with transition redshift @xmath69 satisfying the observational data .
figure 2 and figure 9 depict the variation of energy density against time in presence of different parameters as presented in the figure for case - i and case - ii respectively .
the figures indicate that , in both the cases energy density is positive valued and decreasing function of time .
it is also approaching zero with the evolution of time .
the variation of bulk viscous pressure and coefficient of bulk viscosity is presented in the figure 3 and figure 10 and figure 4 and figure 11 for case - i and case - ii respectively . from the figures
, we noticed that bulk viscous pressure and coefficient of bulk viscosity are negative and positive valued function of time respectively . also , the coefficient of bulk viscosity decreases with the evolution of time and maintains a constant rate after @xmath135
the variation of energy conditions against time for both the cases is presented from figure 5 to figure 7 and figure 12 to figure 14 respectively . in both the cases , energy conditions ( wec , sec , dec ) are satisfied . all the physical parameters presented in both
the cases follow the same quantitative behaviour as that of observational data .
in this article , we have investigated the lrs bianchi type i cosmological model in presence of bulk viscosity in the framework of @xmath0 gravity . according to the choice of @xmath0 we have presented two cosmological models .
the exact solutions of the modified einstein s field equations are obtained under the choice of deceleration parameter of the form ( 19 ) .
the observations of both the models are as follows : * both the models presented here are accelerating and the expanding universe models follow an exponential expansion .
* energy density and coefficient of bulk viscosity are positive valued and decreasing function of time in both the cases and also @xmath136 when @xmath137 . *
bulk viscosity pressure @xmath138 is negative valued in both the cases .
* energy conditions ( sec , wec , dec ) are satisfied for both cases .
* jerk parameter and state - finder trajectory in the @xmath120 plane are close to @xmath85cdm model .
we are very indebted to the editor and the anonymous referee for illuminating suggestions that have significantly improved our paper in terms of research quality as well as presentation .
0 p. a. r. ade _ et al .
_ ( planck collaboration ) , planck 2015 results .
cosmological parameters , _ a_&_a _ ,
* 594 * ( 2016 ) a13 , arxiv:1502.01589 .
( boss collaboration ) , the clustering of galaxies in the completed sdss - iii baryon oscillation spectroscopic survey : cosmological analysis of the dr12 galaxy sample , arxiv:1607.03155 .
( actpol collaboration ) , the atacama cosmology telescope : cmb polarization at @xmath139 , _ jcap _ , * 1410 * ( 2014 ) 10 , arxiv:1405.5524 .
m. garnavich , et al . , constraints on cosmological models from hubble space telescope observations of high - z supernovae .
, _ astrophys .
j. , * 493 * _ ( 1998 ) l53 .
a. g. riess _
et al . _ , observational evidence from supernovae for an accelerating universe and cosmological constant . , _ astron .
* 116 * ( 1998 ) 1009 .
s. perlmutter _
et al . _ ,
measurements of omega and lambda from 42 high - redshift supernovae , _ astrophys . j. _ ,
* 517 * ( 1999 ) 565 .
p. m. garnavich , _ et al .
_ , supernovae limits on the cosmic equation of state , _ astrophys . j._,*509 * ( 1998 ) 74 - 79 .
s. perlmutter , _ et al .
_ , measurements of the cosmological parameters omega and lambda from the first 7 supernovae at @xmath140 , _ astrophys . j. _ * 483 * ( 1997 ) 565 .
p. s. letelier , string cosmologies .
d _ , * 28 * ( 1983 ) 2414 .
d. n. spergel _
_ , first year wilkinson microwave anisotropy probe ( wmap ) observations : determination of cosmological parameters , _ astrophys .
_ , * 148 * ( 2003 ) 175 - 194 , arxiv : astro - ph/0302209
. d. n. spergel _
et al . _ ,
wilkinson microwave anisotropy probe ( wmap ) three years results : implications for cosmology .
, _ astrophys .
_ , * 170 * ( 2007 ) 377 , arxiv : astro - ph/0603449 . m. tegmark _
et al . _ ,
cosmological parameters from sdss and wmap .
d _ , * 69 * ( 2004 ) 103501 .
s. f. daniel et al .
, large scale structure as a probe of gravitational slip , _ phys . rev .
* 77 * ( 2008 ) 103513 , arxiv:0802.1068 .
r. r. caldwell , m. doran , bianchi type - ix magnetized dark energy model in saez - ballestar theory of gravitation , _ phys .
d _ , * 69 * ( 2004 ) 103517 .
z. y. huang__et al . _
_ , holographic explanation of wide - angle power correlation suppression in the cosmic microwave background radiation , _
j. cosmol .
_ , * 05 * ( 2006 ) 013 .
r. r. caldwell , _ et al .
_ , sudden gravitational transition , _ phys .
d _ , * 73 * ( 2006 ) 023513 .
j. t. nielsen , a. guffanti , s. sarkar , marginal evidence for cosmic acceleration from type ia supernovae , _ sci .
_ , * 6 * ( 2016 ) 35596 , arxiv:1506.01354 .
h. shariff , x. jiao , r. trotta , d.a .
van dyk , bahamas : new snia analysis reveals inconsistencies with standard cosmology , _ astrophys .
* 827 * ( 2016 ) 1 , arxiv:1510.05954 .
p. a. r. ade _ et al .
_ ( plank collabaration ) planck 2013 results
cosmological parameters , arxiv:1303.5076 .
s. capozziello , curvature quintessence .
d _ , * 11 * , ( 2002 ) 483 - 492 .
s. m. caroll , v. duvvuri , m. trodden , m. s. turner , is cosmic speed - up due to new gravitational physics ?
, _ phys . rev .
d _ , * 70 * ( 2004 ) 043528 .
a. d. dolgov , m. kawasaki , can modified gravity explain accelerated cosmic expansion ?
b _ , * 573 * ( 2003 ) 1 .
s. nojiri , s. d. odintsov , modified gravity with negative and positive powers of curvature : unification of inflation and cosmic acceleration , _ phys .
* 68 * ( 2003 ) 123512 .
s. nojiri , s. d. odintsov , the minimal curvature of the universe in modified gravity and conformed anomaly resolution of the instabilities , _ mod .
a _ , * 19 * ( 2004 ) 627 - 638 .
m. c. b. abdalla , s. nojiri , s. d. odintsov , consistent modified gravity : dark energy , acceleration and the absence of cosmic doomsday , _ class .
* 22 * ( 2005 ) l35 , arxiv : hep - th/0409177 . o. mena , j. santiago , j. weller , constraining inverse - curvature gravity with supernovae , _ phys .
_ * 96 * ( 2006 ) 041103 , arxiv : astro - ph/0510453
. k. bamba , s. nojiri , s. d. odintsov , the future of the universe in modified gravitational theories : approaching a finite - time future singularity , _ j. cosmol
. astropart .
_ * 0810 * ( 2008 ) 045 .
o. betrolami , c. g. boehmer , t. harko , f. s. n. lobo , extra force in @xmath8 modified theories of gravity , _ phys .
d _ , * 75 * ( 2007 ) 104016 .
s. nojiri , s. d. odintsov , unified cosmic history in modified gravity : from @xmath141 theory to lorentz non - invariant models , _ phys .
_ , * 505 * , ( 2011 ) 59 - 144 , arxiv:1011.0544
. t. harko , f. s. n. lobo , s. nojiri , s. d. odintsov , @xmath0 gravity , _ phys .
d _ , * 84 * ( 2011 ) 024020 .
r. myrzakulov , l. sebastiani , s. vagnozzi , inflation in @xmath13-theories and mimetic gravity scenario , _ eur .
phys . j. c _ , * 75 * ( 2015 ) 444 .
l sebastiani , r myrzakulov , @xmath141-gravity and inflation , _ int . j. geom .
_ , * 12 * , ( 2015 ) 1530003 , arxiv:1506.05330
. m. j. s. houndjo , reconstruction of @xmath0 gravity describing matter dominated and accelerated phases , _ int . j. mod
d _ , * 21 * ( 2012 ) 1250003 .
p. k. sahoo , b. mishra , g. c. reddy , axially symmetric cosmological model in @xmath0 gravity .
j. plus _ , * 129 * ( 2014 ) 49 .
p. k. sahoo , b. mishra , kaluza - klein dark energy model in the form of wet dark fluid in @xmath0 gravity _ can .
j. phys . _
* 92 * ( 2014 ) 1062 - 1067 .
moraes , r.a.c .
correa , g. ribeiro , the starobinsky model within the @xmath0 formalism as a cosmological model , arxiv:1701.01027 .
w. zimdahl _
et al . _ , cosmic antifriction and accelerated expansion . , _ phys .
d _ , * 64 * ( 2001 ) 063501 .
c. w. misner , transport processes in the primordial fireball , _ nature _ , * 214 * ( 1967 ) 40 - 41 . c. w. misner , the isotropy of the universe , _ astrophys . j. _ , * 151 * ( 1968 ) 431 - 457 .
z. klimek , entropy per particle in the early bianchi type - i universe _ il nuovo cimento b _
, * 35 * ( 1976 ) 249 - 258 .
l. p. chimento , s. j. alejandro , d. pavon , enlarged quintessence cosmology _ phys .
d _ , * 62 * ( 2000 ) 063508 . g. p. singh , a.y.kale , anisotropic bulk viscous cosmological models with particle creation , _ astrophysics and space science _ , * 331 * ( 2011 ) 207 - 219 .
a. k. yadav , a. pradhan , a. k. singh , bulk viscous lrs bianchi - i universe with variable @xmath142 and decaying @xmath85 , _ astrophysics and space science _ , * 337 * ( 2012 ) 379 - 385 .
m. jamil , d momeni , m. raza , r. myrzakulov , reconstruction of some cosmological models in @xmath0 cosmology , _ eur .
j. c _ , * 72 * ( 2012 ) 1999 . c. p. singh , s. kumar , a. pradhan , early viscous universe with variable gravitational and cosmological constants , _ class .
quantum grav . _
* 24 * ( 2007 ) 455 .
m. g. hu , x. h. meng , bulk viscous cosmology : statefinder and entropy , _ phys .
b _ * 635 * ( 2006 ) 186 - 194 .
a. d. linde , inflationary cosmology , _ lect .
notes phys . _
* 738 * ( 2008 ) 1 - 54 .
p. a. r. ade , _ et al .
_ ( planck collaboration ) , planck 2013 results .
cosmological parameters , _ astron .
astrophys . _
* 571 * ( 2014 ) a16 .
v. a. belinskii , i. m. khalatnikov , influence of viscosity on the character of cosmological evolution , _ sov .
* 42 * ( 1975 ) 205 .
r. bali , j. p. singh , bulk viscous bianchi type i cosmological models with time - dependent cosmological term @xmath85 , _ int .
* 47 * ( 2008 ) 3288 - 3297 . c. p. singh , s. kumar , viscous fluid cosmology in bianchi type - i space - time , _ int .
* 48 * ( 2009 ) 925 - 936 .
j. c. fabris , s. v. b. goncalves , r. s. rebeiro , bulk viscosity driving the acceleration of the universe , _ gen .
* 38 * ( 2006 ) 495 - 506 .
b. saha , nonlinear spinor field in bianchi type - i universe filled with viscous fluid : numerical solutions , _ astrophysics and space science _
* 312 * ( 2007 ) 3 - 11 .
m. sharif , m. zubair , dynamics of a magnetized bianchi @xmath143 universe with anisotropic fluid , _ astrophysics and space science _ * 339 * ( 2012 ) 45 - 51 .
m. sharif , m. zubair , energy conditions constraints and stability of power law solutions in @xmath0 gravity , _ j. phys .
* 82 * ( 2012 )
014002 . m. sharif , m. zubair , study of bianchi i anisotropic model in @xmath0 gravity , _ astrophysics and space science _
, * 349 * ( 2014 ) 457 - 465 . c. l. bennet _
et al . _ ,
first - year wilkinson microwave anisotropy probe ( wmap ) observations : preliminary maps and basic results , _ astrophys . j. suppl . _ * 148 * ( 2003 ) 1 , arxiv : astro - ph/0302207 .
s. capozziello , o. farooq , o. luongo , b. ratra , cosmographic bounds on the cosmological deceleration - acceleration transition redshift in @xmath8 gravity , _ phys .
d _ , * 90 * ( 2014 ) 044016 , arxiv:1403.1421 .
s. capozziello , o. luongo , e. n. saridakis , transition redshift in @xmath12 cosmology and observational constraints , _ phys .
d _ , * 91 * , ( 2015 ) 124037 , arxiv:1503.02832 .
omer farooq , foram madiyar , sara crandall , bharat ratra , hubble parameter measurement constraints on the redshift of the deceleration - acceleration transition , dynamical dark energy , and space curvature , _ astrophys .
* 835 * ( 2017 ) 26 , arxiv:1607.03537 .
moraes , g. ribeiro , r.a.c .
correa , a transition from a decelerated to an accelerated phase of the universe expansion from the simplest non - trivial polynomial function of @xmath2 in the @xmath0 formalism , _ astrophys space sci .
_ , * 361 * ( 2016 ) 227 , arxiv:1602.07159
. r. m. wald , _ general relativity _ , ( chicago , il : university of chicago press , 1984 )
. s. m. carroll , m. hoffman , m. trodden , can the dark energy equation - of - state parameter w be less than 1 ? , _ phys .
, * 68 * , ( 2003 ) 023509 , arxiv : astro - ph/0301273 . k. lake , _ class . quantum gravity _ * 21 * , ( 2004 ) li29 , arxiv : gr - qc/0407107 . m. zubair , s. waheed , y.ahmad , static spherically symmetric wormholes in @xmath0 gravity , _ eur . phys .
j. c _ * 76 * ( 2016 ) 444 .
m. sharif , s. rani , r. myrzakulov , analysis of @xmath0 gravity models through energy conditions , _ eur .
j. plus _ * 128 * ( 2013 ) 123 .
m. a. h. maccallum , a class of homogeneous cosmological models iii : asymptotic behaviour , _ commun .
_ , * 20 * ( 1971 ) 57 - 84 .
v. sahni , exploring dark energy using the statefinder , astro - ph/0211084 .
m. visser , cosmography : cosmology without the einstein equations , _ gen .
_ , * 37 * ( 2005 ) 1541 - 1548 .
d. rapetti , s. w. allen , m. a. amin , r. d. blandford , a kinematical approach to dark energy studies , _ mon . not .
r. astron .
_ , * 375 * ( 2007 ) 1510 - 1520 , arxiv : astro - ph/0605683
. a. g. riess _
et al . _ ,
type ia supernova discoveries at @xmath144 from the hubble space telescope : evidence for past deceleration and constraints on dark energy evolution , _ astrophys . j. _ ,
* 607 * ( 2004 ) 665 - 687 .
p. astier _
_ , the supernova legacy survey : measurement of @xmath145 and @xmath83 from the first year data set , _ astron
. astrophys .
_ , * 447 * ( 2006 ) 31 - 48 .
v. sahni , t. d. saini , a. a. starobinsky , u. alam , statefinder a new geometrical diagnostic of dark energy , _ jetp lett .
_ * 77 * , ( 2003 ) 201 - 206 . c. j. feng ,
statefinder diagnosis for ricci dark energy , _ phys .
b _ * 670 * , ( 2008 ) 231 - 234 . | the objective of this work enclosed with the study of spatially homogeneous anisotropic bianchi type - i universe in @xmath0 gravity ( where @xmath1 is the ricci scalar and @xmath2 is the trace of stress energy momentum tensor ) in two different cases viz . @xmath3 and @xmath4 with bulk viscosity matter content . in this study , we consider a time varying deceleration parameter , which generates an accelerating universe to obtain the exact solution of the field equations .
the physical and kinematical properties of both the models are discussed in detail for the future evolution of the universe .
we have explored the nature of wec , dec , sec and energy density for both the cases .
we have found that both the models , with bulk viscosity matter component , show an acceleration of the universe .
we have also shown that the cosmic jerk parameter is compatible with the three kinematical data sets . |
giant planets are thought to form either via core - accretion @xcite or gravitational instability scenario @xcite . to get a handle on which formation scenario led to an observed gas giant
, the post - formation entropy of the planet was initially thought to distinguish between the two cases @xcite .
traditionally , planets formed by core accretion were thought to have a low luminosity and entropy ( @xmath09.5 @xmath1/baryon ) corresponding to the so called cold - start " scenario whereas gravitational instability was thought to lead to giant planets having a high luminosity and entropy the hot - start " scenario ( @xmath29.5 @xmath1/baryon , @xcite ) .
recent studies , however , pointed out that the situation is more complex .
the entropy of the planets is affected by whether the accretion of gas onto the planet happens through a supercritical shock front , as indicated by one - dimensional spherically symmetric models @xcite .
if the gas forming the planet passes through such a entropy - reducing shock where a significant part of the accretion luminosity can be radiated , the planet itself will have a low entropy , consistent with the cold - start " scenario , regardless which formation mechanism builds the gas giant @xcite .
moreover , the mass of the solid planetary core can alter the post - formation entropy of the planet as well , with higher mass cores leading to hotter planets @xcite .
finally , @xcite showed that the presence of a circumplanetary disk around the planet will funnel hot gas to the planet that can inflate the outer layers of the gas giant , enhancing the planet s entropy .
observations of directly imaged planets allow to measure the luminosities of young ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and recently also of still forming embedded planets ( e.g. , * ? ? ?
* ; * ? ? ?
this makes it possible to estimate their entropies @xcite and conclude whether they are consistent with the `` cold - start '' or hot - start " scenarios .
a handful of directly imaged gas giants is available for study today , and most seem to be luminous and consistent @xcite with the hot - start " scenario ( which could in principle also be an observational bias as fainter planets are more difficult to detect ) .
however , gravitational instability as a formation mechanism appears unlikely in several of these cases due to various factors , such as the rather small semi - major axis or a rather low circumstellar disk mass @xcite .
furthermore , the luminosity estimations of forming embedded planets from direct imaging observations can be contaminated by the luminosity of the circumplanetary disk or accretion @xcite , enhancing the observed overall luminosity , that can make the planet look like a hot - start " case .
the observation that the currently known directly observed planets seem to be consistent with a hot - start " scenario , but some of them may formed via core accretion ( e.g. , @xmath3 pic b , @xcite ) indicates that solely the low versus high entropy state of an observed gas giant can not conclusively distinguish between the two planet formation scenario .
this is partially due to the lack of theoretical studies of the thermodynamics of giant planet formation which predict the post - formation entropy based on multi - dimensional , radiation - hydrodynamical simulations . in this paper
we therefore present a thermodynamical study based on 3d , radiative hydrodynamic simulations of forming gas giants with various masses embedded in circumstellar disks .
the planets form circumplanetary disks or circumplanetary envelopes around them depending on the gas temperature in the planet vicinity @xcite .
as described e.g. in @xcite , the accretion of the gas happens from the vertical direction through the planetary gap .
this is because the top layers of the circumstellar disk try to close the gap opened by the giant planet , and , as gas enters the gap , it falls nearly freely onto the circumplanetary disk s surface and onto the polar regions of the protoplanet .
the gas shocks at this surface , and then becomes the part of the disk , where it eventually spirals down to the planet . in this work
we study the change of important thermodynamic quantities like the entropy or ionization state of the gas as it passes through the shock front .
our study is based on three - dimensional , radiative , grid - based hydrodynamic simulations with the jupiter code @xcite , developed by f. masset & j. szulgyi .
this code is based on a shock - capturing godunov s method using riemann - solvers and has nested meshes , which allow to zoom onto the planet s vicinity with high resolution .
the radiative module is based on the flux limited diffusion approximation ( e.g. * ? ? ?
* ) , and calculates temperatures where viscous heating and radiative cooling is included .
the opacity of the gas and dust is taken into account through the @xcite opacity table .
this means that despite of the purely gas hydrodynamic simulations , the impact of the dust on the temperature is included through the dust opacities and a fixed dust - to - gas ratio of 1% .
more details about the radiative code and the simulations is given in @xcite and in a future paper for the 3 - 10 @xmath4 gas giants .
here we only mention the most important characteristics .
the coordinate system is spherical , centered onto the star , and covers the entire circumstellar disk radially between 2.0 and 12.4 au with an initial surface density @xmath5 with @xmath6 ( @xmath7 being the semi - major axis , 5.2 au ) .
this density was chosen to be close to the minimum mass solar nebula ( @xcite ) .
only the lower half of the circumstellar ( and circumplanetary ) disks are simulated , assuming symmetry to the disk mid - plane .
the equation of state ( hereafter , eos ) in the code is the ideal gas eos with an adiabatic index @xmath8 equal to 1.43 , such that @xmath9 , where @xmath10 is the internal energy of the gas ( @xmath11 ) , and @xmath12 is the pressure .
the mean molecular weight is 2.3 , corresponding to the solar value . due to the fixed adiabatic exponent and
mean molecular weight , the ionization and dissociation of hydrogen and helium is not taken into account .
the lack of these mechanisms could alter the predicted temperatures , entropies , and compositions , therefore our computations are only estimates found with an idealized eos . due to the extensive computation time needed for the high - resolution radiative simulations presented here , and due to the limitations of the solver
, we had to use in this work an ideal eos .
we performed simulations with 1 , 3 , 5 , and 10 @xmath4 planets for @xmath13 orbits until steady state has been reached .
a constant kinematic viscosity was applied with value of @xmath14 , where @xmath7 is the semi - major axis , @xmath15is the orbital frequency .
this viscosity corresponds to a value of @xmath16 at 5.2 au . due to the nested meshes centered on the planet , the resolution changes mesh by mesh , with each level of refinement doubling the resolution
this way the highest level of resolution ( reached on level 6 ) was @xmath17au@xmath18 where @xmath19 is the diameter of jupiter . in the simulations the planet is treated as a point - mass in the corner of 8 cells . in other words
there is only a gravitational potential well , no sphere is modeled for the giant planet . the smoothing lengths were @xmath20 , @xmath21 , @xmath21 au , and @xmath22 au for the 1 , 3 , 5 , and 10 @xmath4 planets , respectively .
as it was shown in @xcite , in the 1 @xmath4 simulation the gas is too hot in the planet s vicinity to collapse into a circumplanetary disk . instead , a pressure supported spherical envelope formed around the planet .
in this case , there is no shock front on the surface of the envelope where the vertical influx hits it .
we show the entropies of this simulation for comparison , in order to distinguish between cases with and without supercritical shock fronts .
to estimate the specific entropy ( @xmath23 ) of the gas , we used four different approaches .
all four give the specific entropy based on the pressure and the temperature from the hydrodynamic simulations .
the first expression we applied is the classical sackur - tetrode entropy formula for @xmath24/he mixture with he mass fraction of 25% @xcite : @xmath25 this expression does not include ionization / dissociation , such as the eos used in the hydrodynamic code .
the left - hand panel of fig .
[ fig::entr ] shows the specific entropy calculated with the sackur - tetrode expression on a vertical slice through the 3 @xmath26 planet , zooming to the planet vicinity .
the circumplanetary disk pops out with blue colors , which has clearly lower entropy than the supersonic vertical gas influx ( red - yellow colors ) which feeds the circumplanetary disk .
the minimum entropy can be found in a spherical envelope around the planet , within the inner circumplanetary disk . in fig .
[ fig::entr ] we also show the same vertical cut of the density and a further zoomed - in temperature map in the middle and the right - hand panels , respectively . on the temperature color - map
one can see with bright yellow color part of the razor - sharp shock layer ( the zeldovich spike ) right above the planet and on the surface of the inner circumplanetary disk .
the shock however , is even more extended on the top layer of the circumplanetary disk , based on maps of the mach number .
the temperature just above and below zeldovich spike is identical , meaning that this shock front is super - critical @xcite for all the different mass planets which are forming circumplanetary disks ( 3 - 10 @xmath4 planets ) . in the case of the 1 jupiter - mass simulation , as it was mentioned in the previous section , the circumplanetary gas is too hot to collapse into a disk , therefore an envelope is found ( see fig . 1 in @xcite ) .
due to the extended envelope , the vertical influx does not hit it with large enough velocity ( i.e. subsonically ) to create a shock front on the surface of the envelope .
simulations are the zeldovich spikes at the supercritical shock front .
the gas undergoes a strong entropy reduction ( @xmath27/baryon ) via radiative cooling while passing through the shock . ]
the second eos we used was the saumon - chabrier - van horn eos developed for interiors of giant planets @xcite using the same he / h mass fraction . the third way to calculate the specific entropies was using the cea gibbs minimizer @xcite
this approach , unlike the sackur - tetrode expression , calculates the specific entropy , mean molecular weight , and adiabatic index with the dissociation and ionization of hydrogen and helium taken into account .
however , for comparison we also did a calculation with this code enforcing the molecularity and neutrality of the hydrogen - helium gas ( only h@xmath28 and he ) .
this corresponds again to a situation close to the eos used in the hydrodynamic simulations , except for the inclusion of the variation of the degrees of freedom with temperature .
[ fig::shockprof_cdp ] shows a 1d vertical profile of the entropy as the gas passes through the shock front right below the planet ( vertical cut through radius=5.2 au , azimuth=0.0 ) . as mentioned in sect .
[ sec : methods ] , only the lower half of the circumstellar disk is simulated , therefore the co - latitute ranges in this figure from 1.562 radians to 1.57079 ( @xmath29 ) , the latter being the midplane value .
the entropies shown were calculated with cea assuming only h@xmath28 and he .
one can observe that the gas lost significant amount of entropy ( @xmath30/baryon ) when passing through the shock of the circumplanetary disk .
the shock front itself is visible as a spike ( the zeldovich spike ) .
it is obvious that in the vertical influx and at the zeldovich spike the entropy is very high ( 17 - 23 @xmath31/baryon ) as we are dealing with low - density preheated gas falling from the top of the gap .
the entropy is minimal at the planet location ( around 13 @xmath1/baryon for each planet ) .
this shows that also in 3d , an accretion shock is found to play a crucial role in regulating the planet s thermodynamic state . comparing the 3 - 10 @xmath4 gas giants
, it can be observed that with increasing mass , the stronger is the shock , and the higher is the radiative entropy reduction .
this is visible from the decreasing value of the minimum of @xmath23 at some distance behind the shock . also on the mid - plane ( right - end of the figure ) the 10 @xmath4 simulation has the lowest entropy . in the case of the 1 @xmath26 planet , where there is only a circumplanetary envelope without any shock ,
the entropy almost steadily decreasing while approaching the planet .
the shock @xmath32 and post - shock @xmath23 entropies ( measured in and immediately after the zeldovich spike , respectively ) for the four different eos can be found in table [ tab::entropy ] .
the motivation to measure the entropy after the spike is to estimate the entropy of the gas which the planet will eventually accrete .
currently we find high , `` hot - start '' like entropies in the ideal gas eos limit .
the saumon - chabrier eos also give very similar entropy values ( within 0.1 @xmath1/baryon ) to the cea values , as expected .
the sackur - tetrode expression gives somewhat higher ( by 0.6 - 0.7 @xmath1/baryon ) specific entropies .
in contrast to this , in the zeldovich spike itself ( and in a small envelope right around the planet s position ) , the entropy values differ significantly when ionization and dissociation is included in cea relative to the neutral case , since here ionization and dissociation should happen . [ cols="<,<,<,<,<,<,<,<,<",options="header " , ] we also estimated with cea the mass fractions of h ( dissociation ) and h+ ( ionization ) ( fig .
[ fig::ion - diss ] ) for the 10 @xmath4 planet .
for all planetary masses the polar shock surface is so hot ( @xmath33 k ) , that the h@xmath34 molecules are dissociated , and in the 10 jupiter case ( @xmath35 k ) hydrogen is even ionized .
the presence of h+ at the shock surface assumes a detectable h-@xmath36 emission from this extended ( in comparison to a typical planetary radius of a few jovian radii ) region .
this typical accretion tracer was recently detected around the planet , lkca15b @xcite .
we calculated the upper limit for h-@xmath36 flux from the shock of the 10 @xmath4 planet using eq . 21 in @xcite and found @xmath37 .
for the planetary accretion , we applied the expression @xmath38 and its scaling with @xmath39 used in @xcite .
the accretion rates to the planet calculated from our simulations are 9.8 , 13.89 , @xmath40 for the 3 , 5 , 10 @xmath4 planets , which indicate @xmath41 4 to 8 @xmath42 for @xmath43 .
this translates to @xmath41 4 to 7 @xmath44 for h-@xmath36 luminosity , an order of magnitude lower than what was measured in lkca15b .
given that our h-@xmath36 line flux upper limit is three orders of magnitude higher than our planet emitted line flux , it is also plausible that the shock of circumplanetary disk accounts for difference found with @xcite .
our results indicate , that by detecting the h-@xmath36 emission , one can not necessarily distinguish whether it tracks the accretion onto the circumplanetary disk or from the accreting planet itself .
only the lower half of the disk is shown , so the mid - plane is at the top of the figures . in this case
, ionization occurs at the shock front , which suggests that there is h-@xmath36 emission .
dissociation only is found in the shock surfaces of 3 and 5 @xmath4 gas giants . ] when observing a forming embedded giant planet via direct imaging , one has to be careful from where the detected luminosity originates . as we see on the temperature color map of fig .
[ fig::entr ] , part of the shock front on the surface of the circumplanetary disk and protoplanet is very luminous due to shock heating .
this part of the shock front extends to @xmath41100 - 250 @xmath45 in diameter for the 3 to 10 @xmath4 planet simulations .
this part is optically thick in our grey approximation , so it is possible that observationally this shock front is the surface the observations detect , at least in some wavelengths , rather than the actual protoplanet below . because the luminosities of forming directly imaged planets are used to distinguish between `` hot - start '' and `` cold - start '' scenarios , and also to estimate the planetary mass , it is important to consider a possible contribution from the circumplanetary disk shock luminosity .
this is in particular the case if it is not possible to distinguish observationally ( spectroscopically ) the origin of the radiation , e.g. , because hard shock radiation first gets reprocessed in the surrounding disk and re - radiated at longer wavelength .
this highlights that observationally , even intrinsically cold planets could look like `` hot - start '' planets during formation if they are surrounded by luminous shocks on the circumplanetary disk surface . in conclusion ,
circumplanetary disk shock surfaces play a key role not only in regulating a planet s post - formation thermodynamic state , but also for the observational appearance of protoplanets during formation .
in this paper we present a study of the thermodynamics found in global three - dimensional radiative hydrodynamical simulations of embedded accreting giant planets of 1 , 3 , 5 , 10 @xmath4 .
as described in detail in @xcite , the accretional gas flow from the circumstellar disk to the planet is the following
. the gas acts to close the gap opened by the planet in the circumstellar disk , especially so in the high co - latitute regions .
gas enters the gap region , and then falls nearly freely in a vertical influx onto the circumplanetary disk and protoplanet .
because this vertical inflow is supersonic ( mach = 6.2 , 8.1 , 10.3 for the 3 , 5 , 10 @xmath4 planets , respectively ) , it shocks on the surface of the circumplanetary disk and on the polar region of the protoplanet before becoming part of the disk and eventually reaching the protoplanet . in this work we showed that the gas undergoes a significant reduction of the specific entropy ( typically more than 3 @xmath1/baryon ) while passing through the shock front that is found to be supercritical .
the vertical influx has a very high entropy which after the shock in the disk mid - plane reaches a minimum value .
we found that the circumplanetary disk consists of gas of significantly lower entropy than the vertical influx and the lowest entropies are found in a spherical small envelope around the planet within the inner parts of the circumplanetary disk .
we conclude that shocks play a key role in regulating the post - formation entropy . because the shock front on the circumplanetary disk is hot , optically thick ( in our grey - approximation ) , and this luminous region is extended ( 100 - 250 @xmath45 )
, it can contribute strongly to the bolometric luminosity of a directly imaged planet if the gas giant is still accreting .
therefore it is important to disentangle the luminosity of the shock front on the upper layer of the circumplanetary disk from the luminosity of the protoplanet itself beneath the shock surface .
our radiative hydrodynamic simulations compute temperatures taking into account radiative cooling with the inclusion of dust opacities .
however , the use of ideal gas eos does not take into account ionization and dissociation .
this can lead to too high temperatures , and means that we can not yet exactly predict the post - formation entropies . to estimate where dissociation and ionization could occur , we used the cea code to determine potential h and h+ regions .
we found that for 3 - 10 @xmath4 planets dissociation occurs in front of the polar shock surface indicating that non - ideal effect could be important .
in the 10 jupiter - mass simulation the shock surface is so hot to also produced h+ , which means that there could be extended @xmath46 emission from this region , which may be detectable , as seen in lkca 15 b @xcite .
shock surfaces on the surface of the circumplanetary disk and the polar region of the protoplanet could therefore also be important for the observational appearance of forming giant planets .
we are thankful to j. owen , t. guillot , g. marleau for useful discussions and to the anonymous referee for improving this work .
j. sz . acknowledges the support from the eth post - doctoral fellowship by the swiss federal institute of technology .
this work has been in part carried out within the frame of the national centre for competence in research planets " supported by the swiss national science foundation . c.m . acknowledges the support from the swiss national science foundation under grant bssgi0@xmath47155816 `` planetsintime '' .
computations have been done on the mnch " machine ( swiss national comp . cent . )
99 bell , k. r. , & lin , d. n. c. 1994 , apj , 427 , 987 bodenheimer , p. , dangelo , g. , lissauer , j. j. et al .
2013 , apj , 770 , 120 boss , a. p. 1997
, science , 276 , 1836 burrows , a. , marley , m. , hubbard , w. b. , et al .
1997 , apj , 491 , 856 commeron , b. , teyssier , r. , audit , e. , et al .
2011 , a&a , 529 , a35 de val - borro , m. , edgar , r. g. , artymowicz , p. , et al .
2006 , mnras , 370 , 529 forgan , d. , & rice , k. 2013 , mnras , 432 , 3168 hayashi , c. , 1981 , progress of theoretical physics supplement , 70 , 35 kraus , a. l. , & ireland , m. j. 2012 , apj , 745 , 5 lagrange , a .- m . , bonnefoy , m. , chauvin , g. , et al .
2010 , science , 329 , 57 marois , c. , macintosh , b. , barman , t. , et al .
2008 , science , 322 , 1348 marleau , g .- d . , & cumming , a. 2014 , mnras , 437 , 1378 marley , m. s. , fortney , j. j. , hubickyj , o. , et al . 2007 , apj , 655 , 541 mcbride , b. , gordon , s. , & reno , m. 1993 , nasa technical memorandum , 4513 mordasini , c. , alibert , y. , klahr , h. , & henning , t. 2012 , a&a , 547 , a111 mordasini , c. 2013 , a&a , 558 , a113 mordasini , c. , mollire , p. , dittkrist , k .- m . , et al .
2015 , int .
j. of astrobiol .
, 14 , 201 owen , j. e. , & menou , k. 2016 , apjl , 819 , l14 pollack , j. b. , hubickyj , o. , bodenheimer , p. , et al . , 1996 , icarus , 124 , 62 quanz , s. p. , amara , a. , meyer , m. r. , et al .
2015 , apj , 807 , 64 sallum , s. , follette , k. b. , eisner , j. a. , et al .
2015 , nature , 527 , 342 saumon , d. , chabrier , g. , & van horn , h. m. 1995 , apjs , 99 , 713 szulgyi , j. , morbidelli , a. , crida , a. , & masset , f. 2014 , apj , 782 , 65 szulgyi , j. , masset , f. , lega , e. , et al .
2016 , mnras , 460 , 2853 vaytet , n. , et al . 2013 , j. of quantitative spectroscopy and radiative transfer , 125 , 105 zhu , z. 2015 , apj , 799 , 16 | the luminosity of young giant planets can inform about their formation and accretion history .
the directly imaged planets detected so far are consistent with the hot - start " scenario of high entropy and luminosity .
if nebular gas passes through a shock front before being accreted into a protoplanet , the entropy can be substantially altered . to investigate this , we present high resolution , 3d radiative hydrodynamic simulations of accreting giant planets .
the accreted gas is found to fall with supersonic speed in the gap from the circumstellar disk s upper layers onto the surface of the circumplanetary disk and polar region of the protoplanet .
there it shocks , creating an extended hot supercritical shock surface .
this shock front is optically thick , therefore , it can conceal the planet s intrinsic luminosity beneath .
the gas in the vertical influx has high entropy which when passing through the shock front decreases significantly while the gas becomes part of the disk and protoplanet .
this shows that circumplanetary disks play a key role in regulating a planet s thermodynamic state .
our simulations furthermore indicate that around the shock surface extended regions of atomic sometimes ionized hydrogen develop .
therefore circumplanetary disk shock surfaces could influence significantly the observational appearance of forming gas - giants .
[ firstpage ] accretion , accretion discs hydrodynamics methods : numerical planets and satellites : formation planet - disc interactions |
in living cells , membrane shape transformation plays a key role in biological functions such as endo / exocytosis and vesicle transports .
cell organelles have specific shapes depending on their functions .
various types of proteins participate in the regulation of these dynamic and static membrane shapes @xcite .
these proteins mainly control local membrane shapes in two ways : hydrophobic insertions ( wedging ) and scaffolding . in the former mechanism ,
a part of the protein , such as an amphipathic @xmath0-helix , is inserted into the lipid bilayer membrane . in the latter mechanism ,
the protein domain has a strong affinity for the lipid polar head groups and adsorbs onto the lipid membrane . a bar ( bin / amphiphysin / rvs ) domain , which consists of a banana - shaped dimer , mainly bends the membrane along the domain axis via scaffolding @xcite .
some of the bar superfamily proteins , such as n - bar proteins , also have hydrophobic insertions .
experimentally , the membrane tubulation and curvature - sensing by various types of bar superfamily proteins have been observed @xcite . objects with rotational symmetry , such as spherical colloids or conical integral proteins inserted perpendicularly to the membrane , generate an isotropic membrane curvature .
conversely , bar domains which are banana - shaped generate an anisotropic curvature @xcite ( amphipathic @xmath0-helices can also yield an anisotropic curvature @xcite ) .
theoretical models have shown that membrane inclusions , such as adsorbed or embedded proteins , or colloids , undergo long - range interactions that are mediated by the curvature elasticity of the membrane @xcite .
it was also shown that anisotropic inclusions experience interactions of longer range than isotropic ones and are able to produce complex aggregates @xcite . to simplify the theoretical calculations ,
membrane inclusions are usually modelled as non - deformable objects with a fixed curved shape @xcite .
owing to their small sizes it is often convenient to treat them as point - like objects @xcite .
although the existence of membrane mediated interactions has been verified experimentally in the case of isotropic inclusions @xcite , there is no direct experimental evidence yet in the case of anisotropic proteins , despite their biological importance .
numerical simulations are therefore essential in this context .
atomic and coarse - grained molecular simulations @xcite have been employed to investigate molecular - scale interactions between bar proteins and lipids .
the scaffold formation @xcite and linear assembly @xcite of bar domains have been demonstrated . to investigate large - scale membrane deformations , a dynamically triangulated membrane model @xcite and meshless membrane models
@xcite have been employed ; consequently , various ( meta)stable vesicle shapes @xcite and the tubule formation dynamics @xcite have been reported . despite these numerous advancements , the assembly of curved protein rods and their membrane shape - remodeling
have not been understood so far . in living cells , more than one type of bar and other proteins cooperatively work to regulate membrane shape . however , in most of the previous reports , rods of a single type are considered in theories and simulations . to our knowledge , only two studies have been reported for the interactions of two types of proteins .
the mixture of inclusions with isotropic and anisotropic curvatures was shown to produce the self - assembly of a neckless of anisotropic inclusions around a domain consisting of a lattice of isotropic inclusions , mimicking the assembly of dynamin proteins around a scaffolded bud @xcite .
the phase segregation of rods with different positive spontaneous curvatures was reported in ref . .
in the present study , we investigate the membrane - mediated interactions between two types of protein rods with opposite spontaneous curvatures , theoretically and numerically .
this corresponds to the situations in which oppositely curved proteins , such as i - bar and other bar proteins , are adsorbed on the same leaflet of a bilayer membrane , or alternatively , two types of positively curved proteins are adsorbed on opposite leaflets . in our theoretical analysis
we use a multi - scale approach .
first we consider the interaction between two rods of possibly different curvatures .
we assume that the rods are separated by a distance larger than their size so that we can treat them as point - like anisotropic inclusions using the green function formalism of refs . .
contrary to previous models @xcite we assume here , in order to model banana - shaped bar domains , that the inclusions fix a curvature in the rod s direction but do not impose any curvature in the orthogonal direction .
we also consider the interaction under nonzero surface tension , contrary to what was done in previous works .
our results show that rods of same curvature self - assemble into long straight structures that attract rods of opposite curvature on their sides .
we then change the scale of our analysis and study the interaction between these macroscopic straight structures .
we find that their interaction depends crucially on membrane tension . in our simulations
an implicit - solvent meshless membrane model @xcite is used to represent a fluid membrane .
banana - shaped proteins , assumed to be strongly adsorbed onto the membrane , are modelled together with the membrane region below them as linear strings of particles with a bending stiffness and a preferred curvature . in order to investigate the membrane - mediated interactions ,
no direct attractive interaction is considered between the rods .
we investigate the interaction and structures produced by a mixture of a large number of rods of two types .
we find that the results of our simulations agree very well with the theoretical predictions . in sec .
[ sec : theory ] , our multi - scale theoretical analysis of the interactions between curved rods adsorbed onto a membrane with bending rigidity and tension is presented . in sec .
[ sec : method ] , the simulation model and method are described . in secs .
[ sec : flat ] and [ sec : ves ] , the assembly of protein rods in flat membranes and vesicles are presented .
the summary and discussion are given in sec .
[ sec : sum ] .
we consider bar - like membrane inclusions shaped as rods of length @xmath1 that are curved in a plane perpendicular to the plane of the membrane . in order to compute the interaction between such inclusions , at separations well larger than @xmath2
, we model them as a point - like inclusions that impose some membrane curvature @xmath3 along the rod direction and no constraint along the orthogonal direction . to derive the interaction between two such rods we use the green function formalism of refs . .
we first discuss the case of a tensionless membrane , then we take into account membrane tension .
we first consider two such rods adsorbed onto a membrane with vanishing tension ( @xmath4 ) .
let @xmath5 be the vector going from the center of rod 1 to the center of rod 2 , and @xmath6 .
we assume that the rods impose curvatures @xmath7 and @xmath8 along the directions @xmath9 and @xmath10 , at angles @xmath11 and @xmath12 with respect @xmath5 and @xmath13 , respectively ( fig .
[ fig : geom ] ) . to preserve the symmetry between the two inclusions , we orient @xmath11 counterclockwise and @xmath12 clockwise .
we call @xmath14 the distance between the rods .
we consider the limit of small membrane deformations .
minimizing the helfrich bending energy of the membrane @xcite with the curvature constraints yields an interaction which is a binary quadratic form in @xmath7 and @xmath8 , as shown in appendix [ sec : apthe ] . we find that unless @xmath15 , or the opposite , the interaction is well approximated by the sole term proportional to @xmath16 up to distances comparable to @xmath2 .
accordingly , in the following we neglect the contributions proportional to @xmath17 and @xmath18 .
the curvature - mediated interaction energy between two rods in the absence of membrane tension , @xmath19 , is thus obtained as @xmath20 .
\label{eq : hint}\end{aligned}\ ] ] here @xmath21 is the bending rigidity of the membrane . we have exhibited the leading order term @xmath22 , which describes the interaction quantitatively well for @xmath23 larger than a few times @xmath2 .
note that this interaction is of longer range than the @xmath24 fluctuation - induced casimir interaction between straight rods and it has a different angular dependence @xcite .
graphs of this interaction are represented in fig .
[ fig : th - id]a , b and c for various orientations and curvatures of the rods .
when one of the rods is parallel to the separation vector to the other rod ( @xmath25 or @xmath26 ) , eq .
( [ eq : hint ] ) gives @xmath27 which is independent of the orientation of the other rod .
in contrast , when the axes of both rods are perpendicular to the separation vector , i.e. , @xmath28 , @xmath29 has the opposite sign and an amplitude three times larger : @xmath30 hence , when two rods are identical , i.e. , @xmath31 , they have a strong attractive interaction at @xmath28 , while they have a weak repulsive interaction at @xmath25 or @xmath26 .
obviously , when they are perpendicular to their separation vector they will attract up to contact since they match exactly and produce less deformation when they superimpose .
when two rods have curvatures of opposite sign , i.e. , @xmath32 , the interactions are opposite . for @xmath25 or @xmath26 ,
the rods have a weak attractive interaction .
rods of opposite curvatures in a tip - to - tip configuration produce no large - scale curvature deformation when they get close , therefore they will attract up to contact
. this will likely be true also for rods not exactly opposite in curvature .
thus , for neighboring rods at equilibrium , we expect identical rods to align preferentially side - to - side and rods of opposite curvature to align tip - to - tip .
the derivation of the interaction in the presence of membrane tension , @xmath33 , is described in appendix [ sec : apthe ] , sec .
[ tensioncase ] . for the same reasons as previously
, we retain in the interaction only the terms proportional to @xmath34 . in the regime of strong tensions ,
i.e. , for @xmath35 , fig .
[ fig : th - id]a , b , and c , show that the interaction @xmath36 displays several similitudes and differences with the tensionless case : 1 . at short distances
the attractive / repulsive behavior is the same as in the tensionless case .
hence , identical rods will bind side - to - side and rods of opposite curvature tip - to - tip , as previously .
2 . the range of the interaction is shorter in the presence of tension : the black curves decay more rapidly that the gray ones .
the interaction in the tip - to - tip configuration is not monotonic in the strong tension case ( see fig . [
fig : th - id]b , black lines ) .
since identical rods attract strongly and align side - to - side , while opposite rods align preferentially tip - to - tip , we expect the following scenario .
identical rods should aggregate into straight rod assemblies , and parallel rod assemblies should either repel one another , if they are made of rods of same curvature , or attract one another side - to - side if they are made with rods of opposite curvatures .
alternate and periodic stripe structures are therefore expected to develop .
we detail in appendix [ sec : apthe2 ] the calculation of the interaction @xmath46 between two coarse - grained parallel rod assemblies of length @xmath47 that are separated by a center - to - center distance @xmath23 .
the result is @xmath48 where @xmath49 is the contribution coming from the extremities of the rods , which is subdominant and which we do not evaluate .
this interaction confirms that parallel rod assemblies of like curvatures repel while rod assemblies of opposite curvatures attract .
the first term , proportional to the length of the rods , vanishes in the absence of tension , i.e. , for @xmath4 .
this can easily be understood as flat membrane patches can fit in between the rod assemblies at no cost in the absence of tension .
however , some interaction will originate from the extremities of the rod assemblies ( @xmath49 ) , similar to that of isolated rods .
in the presence of tension , the energy per unit length will dominate and produce the attraction of rod assemblies of opposite curvature .
we see then two properties : ( 1 ) the interaction is short - ranged , decaying over @xmath37 , ( 2 ) the larger the tension the stronger the interaction .
since the details of the meshless membrane model and protein rods are described in ref . and refs .
, respectively , we briefly describe the model here . a fluid membrane is represented by a self - assembled one - layer sheet of @xmath50 particles .
the position and orientational vectors of the @xmath51-th particle are @xmath52 and @xmath53 , respectively .
the membrane particles interact with each other via a potential @xmath54 .
the potential @xmath55 is an excluded volume interaction with a diameter @xmath56 for all pairs of particles .
solvent is implicitly accounted for by an effective attractive potential @xmath57 .
the bending and tilt potentials are given by @xmath58 and @xmath59 w_{\rm { cv}}(r_{i , j})$ ] , respectively , where @xmath60 , @xmath61 , @xmath62 , @xmath63 is a weight function , and @xmath64 denotes the thermal energy .
the spontaneous curvature @xmath65 of the membrane is given by @xmath66 .
@xcite in this study , @xmath67 and @xmath68 except for the membrane particles belonging to the protein rods .
a protein rod is modeled as a linear chain of @xmath69 membrane particles .
we use @xmath70 , which corresponds to the typical aspect ratio of the bar domains .
the bar domain width is approximately @xmath71 nm , and the length ranges from @xmath72 to @xmath73 nm .
@xcite two types of protein rods , called rods 1 and 2 , are used . as detailed below , rods 1 and 2 have positive and negative spontaneous curvatures @xmath74 and @xmath75 along the rod axis , respectively , and both rods have no spontaneous curvature perpendicular to the rod axis .
hereafter , we call the membrane particles forming a protein rod _ protein particles_. the protein particles in each protein rod are further connected by a bond potential @xmath76 where @xmath77 , @xmath78 , and @xmath79 are used .
the bending potential is given by @xmath80 , where @xmath81 and @xmath82 for the rod 1 and 2 , respectively .
for bonded pairs of protein particles , we use four times larger values of @xmath83 and @xmath84 , together with the corresponding spontaneous curvature , to prevent the rod from bending tangentially in the membrane .
we employ the parameter sets used in ref . .
the membrane has mechanical properties that are typical for lipid membranes : bending rigidity @xmath85 , area of the tensionless membrane per particle @xmath86 , area compression modulus @xmath87 , and edge line tension @xmath88 .
this edge tension @xmath89 is sufficiently large to prevent membrane rupture in this study @xcite .
molecular dynamics with a langevin thermostat is employed @xcite . in the following
, the results are displayed with the rod length @xmath90 for the length unit , @xmath64 for the energy unit , and @xmath91 for the time unit , where @xmath92 is the diffusion coefficient of the membrane particles in the tensionless membrane @xcite .
for the rod assembly on a flat membrane , the @xmath93 ensemble with periodic boundary conditions is used .
the projected area @xmath94 is fluctuated for a constant surface tension @xmath95 while maintaining the aspect ratio @xmath96 @xcite .
the rod-1 curvature is fixed as @xmath97 , which can induce membrane tubulation with a circumference of @xmath98 for the present bending elastic constants .
flat membranes with @xmath99 are investigated with various values of @xmath100 .
the density of the rods 1 and 2 are set to @xmath101 and @xmath102 , where @xmath103 and @xmath104 are the numbers of the rods 1 and 2 , respectively . for the rod assembly on a vesicle
, the @xmath105 ensemble is used at @xmath106 and @xmath107 .
the radius of the vesicle is @xmath108 in the absence of the rods . for the annealing simulations ,
the rod curvatures are changed from @xmath109 to @xmath97 while maintaining the ratio @xmath100 with annealing rate @xmath110 and subsequently the vesicle is equilibrated for @xmath111 . for each simulation condition , the reproducibility is confirmed at least from four different initial conformations .
( a ) the ratio of the rod curvatures , @xmath112 . ( b ) @xmath113 .
( c ) @xmath114 .
( d ) @xmath115 .
the red and blue particles represent the segments of the rod 1 and 2 , respectively .
the gray particles represent membrane particles . ] and
@xmath4 . ]
first , we describe the rod assemblies on the tensionless flat membrane ( @xmath4 ) .
figure [ fig : snap_cs ] shows typical snapshots . for zero rod - curvature of the rods 2 ( @xmath116 ) , the rods 1 form disk - shaped tubules like in the absence of rods 2 .
the rods 2 are distributed on the flat region of the membrane .
small tubules have a semicircular - disk shape . in contract ,
large tubules have a disk shape ( like a mussel shell ) connected to the flat membrane by a narrow cylindrical neck ( see the tubules in fig . [
fig : snap_cs]a ) .
this assembly of the rods 1 agrees with the theoretical prediction for the point - like inclusions in sec .
[ sec : tpoint ] ( identical curved rods attract and align side - to - side ) .
identical rods have an attractive interaction perpendicular to the rod axes .
the neighboring rods 1 contact each other and subsequently many rods 1 assemble and form a straight one - dimensional band curved along the short axis of the assembly ( parallel to the rod axis ) . eventually , long band assemblies bend along their long axis and form a tubule , as this reduces the bending energy cost around the assembly .
the rods 2 , with @xmath117 , exhibit little interactions with the rods 1 .
this also agrees with the prediction of eq .
( [ eq : hint ] ) . as @xmath118 increases ,
the necks of the large tubules remain open . for
@xmath113 , the rod-1 assemblies exhibit wide walls or hill shapes ( see fig .
[ fig : snap_cs]b ) . the rods 2 are concentrated around the foot of these walls and stabilize their negative curvature . with further increasing @xmath118
, the rods 2 more clearly assemble at both sides of the rod-1 assemblies and form long straight bumps [ see fig . [
fig : snap_cs]c and the corresponding movie provided in esi ( movie 1 ) ] .
these structures recall the shape of a centipede or myriapod : the rod-1 straight assembly looks like the body of a centipede and the rods 2 on both sides are like many legs .
this attraction of oppositely curved rods along their axis also agrees with the theoretical prediction for the point - like inclusions in sec .
[ sec : tpoint ] ( oppositely curved rods attract and align tip - to - tip ) .
the straight assemblies are also stable at larger @xmath118 .
however , at @xmath119 , straight assemblies of opposite curvature have attractive interaction in the lateral direction ( parallel to the axis of each rod ) and form a periodic bump structure ( see fig .
[ fig : snap_cs]d ) .
each rod-1 ( -2 ) straight assembly is connected to two rod-2 ( -1 ) straight assemblies on both lateral sides .
we call this assembly of the rods 1 and 2 a _ stripe assembly _ , since it forms periodic bands . this stripe assembly is metastable but has a longer life time than our simulation periods at @xmath115 . even when the stripe assembly is set as an initial conformation , the rod-1 straight assemblies separate to form isolated bumps at @xmath120 . and
( b ) mean vertical rod span @xmath121 for @xmath114 , @xmath122 , and @xmath123 at @xmath124 and @xmath125 . at @xmath126 and @xmath4 ,
the same data are used as in fig .
[ fig : snap_cn5 ] . ] , @xmath122 , and @xmath123 at @xmath4 .
( b ) @xmath124 , @xmath127 , @xmath125 , @xmath128 , and @xmath129 at @xmath114 .
the same data are used as in fig .
[ fig : ar ] for the corresponding parameter sets . ] for ( a ) @xmath130 and ( b ) @xmath131 . ] and the ratio of the rod curvatures @xmath118 . ] figure [ fig : snap_cn5 ] shows the assembly dynamics at @xmath126 [ the corresponding movie is provided in esi ( movie 2 ) ] .
first , short straight assemblies are formed .
as the tips of the bumps approach closely , they fuse into one large straight assembly . in contrast , as they approach laterally , they form stripe assembly of short straight assemblies . as a long bump approaches , the stripe assemblies are disassembled and reassemble into longer assemblies ( see the upper region of the three right snapshots in fig . [
fig : snap_cn5 ] ) . note that some rods of type 2 also contact the rod-1 assemblies at the tips of the assemblies ( see snapshots at @xmath132 and @xmath133 in fig .
[ fig : snap_cn5 ] ) , so that an attractive interaction exists between the rods of types 1 and 2 even when the rod-1 axis is perpendicular to the separation vector .
this attraction is in good agreement with the prediction of eq .
( [ eq : hint ] ) with @xmath25 and @xmath134 .
as the rods assemble , the projected membrane area @xmath135 decreases while the mean vertical rod span @xmath121 and the mean cluster size @xmath136 of rods 1 increase ( see figs . [ fig : ar ] and [ fig : cl ] ) . at @xmath137 , the rod-1 cluster formation becomes slower than at @xmath138 , since the stripe assembly suppresses the fusion between the rod-1 assemblies .
the vertical span is calculated from the height variance of all rod segments as @xmath139 where @xmath140 .
a rod is considered to belong to a cluster when the distance between the centers of mass of the rod and one of the rods in the cluster is less than @xmath141 . for the rods 1 ,
the mean cluster size is given by @xmath142 with @xmath143 where @xmath144 is the number of clusters of size @xmath145 . at small positive surface tensions ( @xmath146 , _
i.e. _ , @xmath147 ) , the bending energy contribution should be dominant in the interactions between the rods as described in sec .
[ sec : theory ] .
our simulation results support this theoretical prediction .
the rod assembly is only slightly modified at @xmath148 ( @xmath149 ) .
in contrast , a large positive surface tension ( @xmath150 ) suppresses both the decrease of the projected area @xmath135 and the protrusion of the rod assemblies ( see fig . [
fig : ar ] ) . since the curvature - energy gain by the rod assembling
is reduced when increasing @xmath95 ( see the theory in sec .
[ sec : theory_tension ] and fig .
[ fig : th - id ] ) , the rod cluster size decreases and more rods remain in an isolated state ( see figs .
[ fig : cl]b and [ fig : snap_f ] ) .
interestingly , the stripe assembly is stabilized for smaller @xmath118 at larger values of @xmath95 , as shown in fig .
[ fig : pd ] . for @xmath114 , the stripe assembly exists for @xmath151 while it does not for @xmath152 ( compare figs . [
fig : snap_cs]c and [ fig : snap_f ] ) .
this agrees well with the theoretical results of sec .
[ sec : tbars ] showing that the attraction between long straight assemblies of opposite curvature scales as @xmath153 and thus increases with membrane tension .
of the rods spontaneous curvatures .
( a)(d ) snapshots .
the membrane particles are represented by a transparent grey sphere .
the rod 1 ( resp .
rod 2 ) segments are represented by a sphere half red and half yellow ( resp .
half blue and half green ) .
the orientation vector @xmath53 lies along the direction from the yellow ( green ) to red ( blue ) hemispheres .
( a ) , ( b ) vesicles formed by annealing to @xmath97 with ( a ) @xmath154 and ( b ) @xmath155 .
( c ) vesicles formed by an increase in @xmath118 from @xmath156 to @xmath123 at @xmath97 .
( d ) vesicles formed by a decrease in @xmath118 from @xmath157 to @xmath158 at @xmath97 .
top and bottom panels show the snapshot of all particles in birds - eye view and a sliced snapshot in front view , respectively .
( e),(f ) dependence of ( e ) the mean radius of gyration @xmath159 and ( f ) mean asphericity @xmath160 of vesicles on @xmath118 .
the data labeled up and down are obtained by an increase and decrease in @xmath118 at @xmath97 .
, title="fig : " ] of the rods spontaneous curvatures .
( a)(d ) snapshots .
the membrane particles are represented by a transparent grey sphere .
the rod 1 ( resp .
rod 2 ) segments are represented by a sphere half red and half yellow ( resp .
half blue and half green ) .
the orientation vector @xmath53 lies along the direction from the yellow ( green ) to red ( blue ) hemispheres .
( a ) , ( b ) vesicles formed by annealing to @xmath97 with ( a ) @xmath154 and ( b ) @xmath155 .
( c ) vesicles formed by an increase in @xmath118 from @xmath156 to @xmath123 at @xmath97 .
( d ) vesicles formed by a decrease in @xmath118 from @xmath157 to @xmath158 at @xmath97 .
top and bottom panels show the snapshot of all particles in birds - eye view and a sliced snapshot in front view , respectively .
( e),(f ) dependence of ( e ) the mean radius of gyration @xmath159 and ( f ) mean asphericity @xmath160 of vesicles on @xmath118 .
the data labeled up and down are obtained by an increase and decrease in @xmath118 at @xmath97 .
, title="fig:",width=302 ]
the rods also assemble on vesicles like on flat membranes . however , the assembly structures are largely modified by the original vesicle curvature and its closed geometry . when a single type of rod exists on the vesicle ,
the rods induce oblate or polyhedral shapes of the vesicle and they assemble at the edges of the polyhedron @xcite . for negative rod curvatures , invaginations into the inside of the vesicle can occur @xcite . here , we set the rod-1 density , @xmath161 , to make the vesicle form a disk shape in the absence of the rods 2 .
we add the rods 2 with the same density and investigate how the vesicle shape is changed upon varying @xmath118 .
figure [ fig : snap_ves]a and b show the snapshots obtained by annealing from @xmath109 to @xmath97 at fixed ratio @xmath118 . for @xmath154 ,
the vesicles form a disk shape with a concave circular region between the center and the rim ; the rods 2 are aligned along radial directions of the disk in this concave region . for @xmath155 , the rods 2 form invaginations into the inside of the vesicle and the two rims of the rod-1 assembly are partially connected by the rod-2 assembly like in the flat membrane . for intermediate values of @xmath118 ,
the obtained vesicle shapes depend on the initial conformations and thermal noise during the annealing : some vesicles have an invagination or partial doubled rims
. to investigate the stability of the disk - shaped and invaginated vesicles , we simulated vesicles with different @xmath118 starting from the annealed conformation at @xmath154 and @xmath157 as initial states .
as @xmath118 increases for disk - shaped vesicles , the center region of the vesicle becomes more spherical .
this shape change reduces the radius of the disk , resulting in a buckling or winding of the disk rim , at @xmath126 and @xmath123 ( see fig .
[ fig : snap_ves]c ) . at @xmath162 and @xmath157 , the rods
2 form a straight dimple on the side of the rod-1 assembly ( as seen in the left region of the sliced snapshot in fig .
[ fig : snap_ves]b ) as well as an invagination .
starting from an invaginated vesicle , the invagination swells to an ellipsoidal shape and the rest of the rods 1 form a circular rim , at @xmath114 ( see fig . [
fig : snap_ves]d ) . at @xmath154 ,
the invagination is removed and a disk - shaped vesicle is obtained [ see the movie is provided in esi ( movie 3 ) ] .
these shape changes can be characterized by the radius of gyration @xmath163 and the shape parameter , asphericity @xmath164 ( see fig . [
fig : snap_ves]e and f ) .
the asphericity is the degree of deviation from a spherical shape and is expressed as @xcite @xmath165 where @xmath166 and @xmath167 are the eigenvalues of the gyration tensor of the vesicle : @xmath168 , where @xmath169 and @xmath170 are the coordinates of the center of mass . for a perfect sphere , @xmath171 , for a thin rod @xmath172 , and for a thin disk , @xmath173 .
starting from a disk - shaped vesicle ( labeled ` up ' in fig .
[ fig : snap_ves]e and f ) , the vesicles have more spherical and less compact shapes than invaginated vesicles so that they take larger @xmath163 and smaller @xmath164 .
thus , the vesicle shapes can be chosen from many metastable conformations by using the hysteresis of @xmath118 variations .
our simulations demonstrate a control method of the vesicle shapes by using hysteresis .
we have studied the assembly structures of binary banana - shaped protein rods .
compared to the situation where only one type of rod is present , the coexistence of two types of rods , with opposite curvatures , induces a greater variety of membrane shapes . here
, using meshless membrane simulations , we demonstrated straight bumps and stripe structures in flat membranes , and buckled rims and ellipsoidal invaginations in vesicles .
more complicated shapes such as vesicles with periodic bumps , like cartridge pleats , can be formed for larger vesicles .
these structure formations can be examined in experiments with two types of bar proteins absorbed on the same side of the membrane or with a single type of protein absorbed on both side of the membrane .
the periodic length of the stripe structure is the sum of the longitudinal lengths of two bar domains ( @xmath174 to @xmath125 nm ) .
it should therefore be observable by electron microscopy . for rods of like curvatures
, our theory predicts a side - to - side attractive interaction and a tip - to - tip repulsion . for rods having curvatures of opposite sign
our theory predicts the reverse , i.e. , a tip - to - tip attraction and a side - to - side repulsion .
our coarse - grained simulations revealed that identical rods assemble in the side - to - side configuration and thus build straight rod assemblies ; these assemblies attract the rods with opposite curvatures in the tip - to - tip configuration thereby inducing a side - to - side alignment of the formed assemblies .
the formation of those alternate assemblies , or stripe structures , therefore agrees well with our theoretical predictions . with increasing membrane tension
, we theoretically found that the attractive interaction between opposite rods in the tip - to - tip configuration ( which corresponds to the orientations involved in the stripe structures ) is increased at short distances and decreased at large distances ( see fig .
[ fig : th - id]b ) .
likewise , our larger - scale analysis of sec .
[ sec : tbars ] predicts a short - range attraction between opposite straight rod assemblies that increases with membrane tension .
these theoretical results explain well the observed increased stability of the stripe structures when the tension is increased and the shorter length of the rod assemblies in the thermalized system .
these theoretical predictions are supported by our simulation results although the theory considers rigid rods with a fixed shape while the rods in the simulation are flexible .
the real proteins are not completely rigid but their stiffness is not measured . the rods flexibility can be considered to result in an effectively smaller rod curvature , not altering the sign of the interactions in the present assemblies .
the protein stiffness , together with the finite size of the rod inclusions , should however be taken into account for quantitative predictions of the assembly conditions .
the effects of the casimir force , induced by the fluctuations of the membrane , remains to be investigated further .
it was ignored in the present work since it is normally sub - dominant for small inclusions imposing large membrane deformations .
this casimir force is known however to induce an effective attraction between long straight rods @xcite .
we have investigated straight rods in our simulations , and found that the latter have too weak casimir interactions to induce rod clusters for the present length ( @xmath70 ) . however , twice longer rods ( @xmath175 ) can form clusters . thus
, the assembly of longer rods can be induced by both casimir and rod - curvature forces . in living cells ,
many types of bar superfamily proteins cooperate to regulate membrane shapes . here , we have studied only the mixture of two types of proteins .
the cooperative effect of mixing more than two proteins is an important topic for further studies .
this work was supported by jsps kakenhi grant number jp25103010 .
in the small deformation regime , the membrane shape can be expressed in the monge gauge by a height function @xmath176 , with @xmath177 a point of the projected plane .
the helfrich elastic energy of the membrane is given , to quadratic order , by @xcite @xmath178 = \frac{\kappa}{2}\int\!d^2r\,h\mathcal{l}h,\end{aligned}\ ] ] with @xmath179 , the operator associated with the hamiltonian , and @xmath180 the coherence length arising from the membrane tension @xmath95 and the membrane bending rigidity @xmath21 .
let us consider a point - like inclusion , placed at @xmath181 , that imposes a membrane curvature @xmath182 along the direction given by the unit vector @xmath9 of angle @xmath11 ( see the definition of the angles in sec .
[ sec : tpoint ] ) , and a second point - like inclusion , placed at @xmath183 , imposing a curvature @xmath184 along the direction given by the unit vector @xmath10 of angle @xmath12 .
we shall discuss later on how to mathematically fix the size " of these inclusions .
the constraints set by the inclusions , for small membrane deformations , read @xmath185 where a comma indicates differentiation . here
@xmath186 is the polar angle of the direction @xmath10 with respect to the @xmath187-axis , i.e. , @xmath188 . minimizing the helfrich energy ( [ eq : hel ] ) with the above constraints
yields the euler lagrange equation : @xmath189 + \lambda_2\big [ \delta_{,xx}(\bm{r}-r\bm{e}_x)\cos^2\theta'_2 \nonumber \\ & + \delta_{,xy}(\bm{r}-r\bm{e}_x)\sin(2\theta'_2 ) + \delta_{,yy}(\bm{r}-r\bm{e}_x)\sin^2\theta'_2 \big ] .
\nonumber\end{aligned}\ ] ] here , @xmath190 is the dirac distribution , and @xmath191 and @xmath192 are lagrange multipliers .
the solution is given by @xmath193 + \lambda_2\big [ g_{,xx}(\bm{r}-r\bm{e}_x)\cos^2\theta'_2 \nonumber \\ & + g_{,xy}(\bm{r}-r\bm{e}_x)\sin(2\theta'_2 ) + g_{,yy}(\bm{r}-r\bm{e}_x)\sin^2\theta'_2 \big ] , \nonumber\end{aligned}\ ] ] where @xmath194 is the green function defined by @xmath195 if @xmath4 , the euler - lagrange operator becomes @xmath196 and the green function is given by @xcite @xmath197 indeed , with @xmath198 ( cylindrical coordinates ) , the most general rotationally symmetric solution of @xmath199 is @xmath200 .
the finiteness of @xmath201 at the origin requires @xmath202 and integrating eq .
( [ eq : green ] ) around @xmath203 yields @xmath204 . the constants @xmath205 and @xmath206 are useful
if one wishes to satisfy some boundary condition for large @xmath207 , however they do not contribute to the fourth derivatives of @xmath201 appearing below , so we can set @xmath208 , which yields eq .
( [ eq : freegreen ] ) . satisfying the constraints ( [ cons1 ] ) and ( [ cons2 ] )
gives a linear set of equations for the lagrange multipliers : @xmath209 where @xmath210 , @xmath211 and @xmath212 is a matrix such that @xmath213 \nonumber\\ & + 2g_{,xyyy}(r\bm{e}_x)\sin\theta_1\sin\theta'_2\sin(\theta_1+\theta'_2 ) \nonumber\\ & + g_{,yyyy}(r\bm{e}_x)\sin^2\theta_1\sin^2\theta'_2\,,\end{aligned}\ ] ] and with @xmath214 obtained from @xmath215 by replacing @xmath216 by @xmath217 , and with @xmath218 obtained from @xmath219 by replacing @xmath11 by @xmath186 .
using eqs .
( [ eq : hel ] ) and ( [ eq : lh ] ) , one obtains exactly @xmath220 we need the fourth derivatives of the green function . from eq .
( [ eq : freegreen ] ) , we obtain @xmath221 it follows that the fourth derivatives of @xmath222 , that appear in @xmath219 and @xmath218 , are singular .
this problem comes from the fact that we use mathematically a point - like constraint .
physically , the inclusions have a characteristic size , here @xmath223 , and we can specify this size by using an upper wavevector cutoff @xmath224 for the membrane deformation in the reciprocal space .
we take numerically @xmath225 as in refs . , since this choice was shown to match the exact calculations in the case of isotropic inclusions .
this gives @xmath226 similarly , @xmath227 we thus obtain @xmath228 where we have switched back to @xmath229 .
note that @xmath214 is symmetric upon the exchange of @xmath11 and @xmath12 as it should be .
using eq .
( [ hbyparts ] ) , we obtain exactly @xmath230 where @xmath231 .
however , since the rods we are actually modeling are not point - like , the above formula is not meaningful for @xmath232 .
we therefore take the leading order of @xmath233 in @xmath234 , which describes the interaction quantitatively well for @xmath23 larger than a few times @xmath235 . replacing @xmath235 by @xmath2 yields the interaction @xmath236 given by eq .
( [ eq : hint ] ) .
note that this result is symmetric upon exchanging @xmath11 and @xmath12 and invariant in changing @xmath11 into @xmath237 or @xmath12 into @xmath238 , as it should be . in the presence of membrane tension , for @xmath239 , the formalism in the previous section
is still valid , provided the full green function is used , i.e. , the green function of @xmath240 defined by eq .
( [ eq : green ] ) .
in particular , eqs .
( [ eq : m11])([hbyparts ] ) still hold . the green function in the presence of tension
is given by @xcite @xmath241,\ ] ] where @xmath242 and @xmath243 ( used below ) are modified bessel functions .
indeed , the most general rotationally symmetric solution of @xmath244 is @xmath245 .
the finiteness of @xmath201 at the origin requires @xmath246 and integrating eq .
( [ eq : green ] ) around @xmath203 yields @xmath247 . discarding the constant term @xmath205 and the term @xmath248 that diverges at infinity
, we obtain eq .
( [ eq : tensiongreen ] ) .
note that eq .
( [ eq : tensiongreen ] ) reduces to eq .
( [ eq : freegreen ] ) for @xmath249 . regularizing the fourth derivatives of the green function as previously , we obtain @xmath250 similarly , @xmath251 using eqs .
( [ eq : m11])([hbyparts ] ) with these elements , we obtain the graphs of @xmath33 shown in sec .
[ sec : theory_tension ] and the corresponding asymptotic behaviors @xmath43 and @xmath45 .
we consider two parallel rod assemblies , one made of rods of curvature @xmath7 and the other one made of rods of curvature @xmath8 .
recall that the rods are curved in the direction perpendicular to the axes of the rod assemblies .
let us define @xmath207 , @xmath252 and @xmath253 in such a way that the curvatures @xmath7 and @xmath8 corresponds to angular variations @xmath254 and @xmath255 over the distance @xmath256 , respectively ( see fig .
[ fig : assemblies ] ) . for the sake of simplicity , we regard the rod assemblies as infinite and homogeneous , i.e. , we neglect extremity effects and the discrete character of the rods .
we fix the distance @xmath257 between the centers of the two rod assemblies and we proceed to calculate the deformation energy stored in the membrane ( taking into account membrane tension ) .
we have , in principle , four degrees of freedom : the heights @xmath258 and @xmath259 of the assemblies with respect to the reference plane that is parallel to the membrane at infinity , and the tilt angle @xmath260 and @xmath261 of the normal to the assemblies relative to the normal to this reference plane .
if we fix these four variables we have the following eight boundary conditions : @xmath262 here the prime indicates differentiation with respect to @xmath187 , the subscript ` @xmath263 ' refers to the position @xmath264 and the subscript ` @xmath265 ' refers to the position @xmath266 .
the total energy of the membrane , per unit length and in the limit of small deformations , is given by the helfrich hamiltonian in the gaussian approximation @xcite : @xmath267,\end{aligned}\ ] ] where @xmath268-\infty ,- d - r]\cup[-d+r , d - r]\cup[d+r,\infty[$ ] .
it has to be minimized with respect to the shape @xmath269 of the membrane in @xmath270 , but also with respect to @xmath258 , @xmath259 , @xmath260 and @xmath261 , which corresponds to requiring that no forces nor torques act on the rod assemblies at equilibrium .
the corresponding conditions can be obtained directly by performing two integration by parts on the first variation of @xmath271 .
we obtain @xmath272\delta h(x)\,dx \nonumber\\ & -\gamma_1\,\delta\alpha_1-\gamma_2\,\delta\alpha_2 -f_1\,\delta h_1-f_2\,\delta h_2,\end{aligned}\ ] ] where we identify @xmath273 , @xmath274 , @xmath275 and @xmath276 as the torques and forces ( per unit length ) acting on the rod assemblies , respectively . requiring them to vanish we obtain a new set of eight boundary conditions , the conditions that must be satisfied at equilibrium : @xmath277 as for the membrane
, it must satisfy the euler - lagrange equation @xmath278 solving this linear , one - dimensional problem , yields the interaction energy per unit length of the rod assemblies @xmath279 which , after setting @xmath256 , @xmath280 , @xmath281 , @xmath282 and multiplying by the length of the rod assemblies , yields eq .
( [ intbars ] ) . | the assembly of banana - shaped rodlike proteins on membranes , and the associated membrane shape transformations , are investigated by analytical theory and coarse - grained simulations .
the membrane - mediated interactions between two banana - shaped inclusions are derived theoretically using a point - like formalism , both for zero surface tension and for finite surface tension . on a larger scale ,
the interactions between assemblies of such rodlike inclusions are determined analytically .
meshless membrane simulations are performed in the presence of a large number of inclusions of two types , corresponding to rods of opposite curvatures , both for flat membranes and vesicles .
rods of the same type aggregate into linear assemblies perpendicular to the rod axis , leading to membrane tubulation .
however , rods of the other type , those of opposite curvature , are attracted to the lateral sides of these assemblies , and stabilize a straight bump structure that prevents tubulation .
when the two types of rods have almost opposite curvatures , the bumps attract one another , forming a stripe structure .
positive surface tension is found to stabilize the stripe formation .
the simulation results agree well with the theoretical predictions . |
the interest in noncentrosymmetric structures , exhibiting superconductivity , has grown up in the past years .
there are only a few examples of this type , which belong to various classes of materials , e.g. an antiferromagnetic heavy fermion system cept@xmath4si @xcite , ferromagnetic uranium compound uir , superconducting under pressure @xcite , or non - magnetic ternary borides li@xmath5pd@xmath4b @xcite and li@xmath5pt@xmath4b @xcite .
the main reason , why these systems are especially attracting , is related to the role of inversion symmetry in electron pairing .
the absence of inversion symmetry may suppress the triplet pairing or mix singlet and triplet symmetry @xcite .
very recently , klimczuk and co - workers @xcite synthesised a new type of intermetallic light - element based compound mg@xmath0ir@xmath1b@xmath2 , exhibiting superconductivity near 5 k. this novel material also belongs to the rare noncentrosymmetric structures , and crystallises in a large and rather complex @xmath6 cell ( space group i-43@xmath7 ) @xcite .
mg@xmath0ir@xmath1b@xmath2 is partly similar to the li@xmath5(pd , pt)@xmath4b system , since these structures contain an alkali metal ( li , mg ) , boron , and heavy transition metal ( pd , pt , ir ) . in this work
we intend to start the discussion on superconductivity mechanism in this unusual compound , analysing the electronic structure and the strength of electron - phonon coupling ( epc ) . assuming the bcs - type behaviour in mg@xmath0ir@xmath1b@xmath2
, we study whether superconductivity is driven by light boron sublattices , like e.g. in mgb@xmath5 , or by heavy transition metal atoms , as suggested for li@xmath5pd@xmath4b @xcite .
electronic structure calculations were performed using the korringa - kohn - rostoker ( kkr ) multiple scattering method @xcite .
the crystal potential was constructed in the framework of the local density approximation ( lda ) , using von barth and hedin formula @xcite for the exchange - correlation part .
for all atoms angular momentum cut - off @xmath8 was set , mesh in the irreducible part of brillouin zone ( irbz ) contained over 200 points .
densities of states ( dos ) were computed using the tetrahedron integration technique , generating over 600 tetrahedrons in irbz . due to the high atomic number of iridium ( @xmath9 )
semi - relativistic calculations were performed , but neglecting the spin - orbit coupling , which is commented at the end of this paper .
the lack of inversion centre is clearly seen e.g. along the main diagonal , where b1 and mg1 atoms break the inversion symmetry . generated by xcrysden @xcite . ]
as far as the crystal structure is concerned , experimental lattice constant and atomic positions @xcite were accounted for the computation ( for clarity also shown in ) .
atoms in the unit cell were surrounded by muffin - tin ( mt ) spheres with following radii : , , ( in atomic units ) , filling about 60% of the cell volume . in the primitive cell of this system ,
45 atoms occupy 7 inequivalent sites , which all are listed in .
the noncentrosymmetricity of this system is important , and can not be regarded as the effect of a lattice distortion , as observed e.g. in uir @xcite .
the inversion symmetry is broken by both boron sublattices ( b1 and b2 ) , as well as mg1 and ir3 sites , thus the crystal has a half of symmetry operations of the cubic group o@xmath10 , i.e. only 24 operations . among of all sublattices ,
the positions of iridium atoms are the closest to have inversion symmetry .
since ir3 occupies @xmath11 sites , with @xmath12 , the full cubic symmetry is restored after shifting this position to @xmath13 .
the analysis of superconducting properties is based on the computed mcmillan - hopfield ( mh ) parameters @xmath14 @xcite which determine electronic part of the electron - phonon interaction , and directly enter the formula for the electron - phonon coupling constant @xmath3 : @xmath15 in equation , @xmath16 is the mh parameter for each nonequivalent atom @xmath17 which is characterised by the atomic mass @xmath18 and averaged squared vibrational frequency @xmath19 .
this equation divides the electron - phonon coupling constant into site - dependent parts , thus it allows to justify which sublattice gives the most important contribution to the total @xmath3 .
calculations of mh parameters were preformed using the rigid muffin tin approximation ( rmta ) @xcite , which gives the following expression for @xmath14 at each atomic site @xmath17 : @xmath20 here , @xmath21 is the angular momentum number , @xmath22 is the partial ( angular - decomposed ) density of states per spin at the fermi energy ( @xmath23 ) , @xmath24 is a regular solution of the radial schrdinger equation , normalised to unity inside the @xmath25 sphere of radius @xmath26 , and @xmath27 is the self - consistent , spherically - symmetric potential .
@xmath28 is the total dos at @xmath23 per spin and per cell . for more detailed discussion of approximations involved in this method , see e.g. @xcite and references therein
the calculated total and atomic densities of states functions are presented in .
as one could expect , electronic structure is dominated by iridium @xmath29 states .
the total dos curve may be characterised as a collection of numerous van hove singularities , reflecting the large number of atoms in the unit cell and various interatomic distances .
the site - decomposed dos are shown in , and their values at the fermi level are gathered in . the total dos at fermi level is about @xmath30 ry@xmath31 per formula unit .
the densities at @xmath23 , calculated per atom , are rather low ( average value per atom : 3.3 ry@xmath31 ) , and generally @xmath23 is located outside of dominating peaks of dos , on a small decreasing slope of ir dos . strong hybridisation of @xmath32 states of ir and @xmath33 states of b and mg is manifested in the separation of bonding and anti - bonding states , and @xmath23 is located in the dos valley ( and ) .
the interatomic distances , listed in , supports the enhanced @xmath33-@xmath32 hybridisation , especially between ir and b atoms ( the smallest distances , 2.1 - 2.2 ) .
@ll|lllll|lllll atom & @xmath34 & @xmath35 & @xmath36 & @xmath37 & @xmath38 & @xmath39 & @xmath16 & @xmath40 & @xmath41 & @xmath42 & @xmath43 + ir1 & 1 & 2.77 & 0.15 & 0.88 & 1.86 & 0.03 & 2.8 & 0.1 & 1.4 & 1.3 & 2.8 + ir2 & 6 & 3.16 & 0.07 & 0.34 & 2.71 & 0.03 & 3.3 & 0.0 & 0.8 & 2.5 & 19.8 + ir3 & 12 & 2.41 & 0.06 & 0.34 & 1.97 & 0.03 & 2.6 & 0.0 & 0.6 & 2.0 & 31.2 + b1 & 4 & 0.40 & 0.03 & 0.35 & 0.01 & 0.00 & 1.2 & 0.0 & 1.2 & 0.0 & 4.8 + b2 & 12 & 0.27 & 0.04 & 0.22 & 0.01 & 0.00 & 0.6 & 0.0 & 0.6 & 0.0 & 7.2 + mg1 & 4 & 0.98 & 0.17 & 0.69 & 0.10 & 0.03 & 0.7 & 0.5 & 0.2 & 0.0 & 2.8 + mg2 & 6 & 0.67 & 0.09 & 0.41 & 0.14 & 0.03 & 0.2 & 0.1 & 0.1 & 0.0 & 1.2 + ir@xmath1b@xmath2 ( @xmath44 ) . the total dos of @xmath45 iridium metal is given for comparison in panel ( d ) . ]
it is also interesting to compare the computed site - decomposed ir densities , to the dos of metallic @xmath45 iridium , which is presented in ( d ) .
the shape of lower part of dos in ir2 and ir3 is quite similar to the case of ir-@xmath45 .
this is probably due to the fact , that the ir2-ir3 and ir3-ir3 coordination , as well as interatomic distances , are very close to the @xmath45 phase ( ir - ir distance is @xmath46 in the aforementioned cases , see ) .
however , the @xmath35 values on ir atoms are much lower , comparing to @xmath45 structure ( @xmath47 6.3 ry@xmath31/spin in ir-@xmath45 ) , being the effect of enhanced hybridisation near @xmath23 .
noteworthy , ir1 has a quite different atomic coordination , with respect to ir2 and ir3 positions , being surrounded practically only by 4 boron atoms ( b1 ) .
it is clearly reflected by the apparently different dos shape below @xmath23 .
electronic dispersion curves near the fermi level are presented in .
the bands located below @xmath48 ry ( partly showed ) are very flat , and form the narrow large dos peaks , seen already in .
conversely , the bands that cross @xmath23 are quite dispersive , which results in the low dos in this energy range .
noteworthy , there is an energy gap along the p - n direction ( parallel to the @xmath49 axis in the reciprocal space ) .
this may suggest some anisotropic transport properties of this compound .
@llllllllcl & ir1 & ir2 & ir3 & b1 & b2 & mg1 & mg2 & atomic position & site + ir1 & 9.1 & 5.9 & 3.9 & 2.1 & 5.0 & 3.1 & 3.6 & ( 0 , 0 , 0 ) & ( 2a ) + ir2 & & 3.8 & 2.7 & 4.5 & 2.2 & 4.0 & 3.1 & ( 0 , 0.25 , 0.5 ) & ( 12d ) + ir3 & & & 2.7 & 2.1 & 2.2 & 4.0 & 2.9 & ( 0.0702 , 0.2525 , 0.2525 ) & ( 24 g ) + b1 & & & & 3.4 & 3.3 & 3.0 & 3.0 & ( 0.3331 , 0.3331 , 0.3331 ) & ( 8c ) + b2 & & & & & 3.3 & 2.6 & 2.5 & ( 0 , 0 , 0.3473 ) & ( 24 g ) + mg1 & & & & & & 5.0 & 3.1 & ( 0.1127 , 0.1127 , 0.1127 ) & ( 8c ) + mg2 & & & & & & & 3.1 & ( 0.1639 , 0.1639 , 0.4140 ) & ( 12e ) + in mg@xmath0ir@xmath1b@xmath2.,scaledwidth=70.0% ] the calculated mcmillan - hopfield factors for all sites , with contributions from each scattering channel ( @xmath50 ) , are presented in .
estimation of the electron - phonon coupling constant @xmath3 , using mh parameters , requires also the knowledge of average phonon frequencies .
for such a large structure , containing 90 atoms in the cubic unit cell , phonon spectra calculations are difficult to be carried out
. nevertheless we can try to investigate the strength of electron - phonon interaction assuming reasonable values of @xmath19 and studying their influence on estimated @xmath3 values .
debye frequencies @xmath51 of monoatomic crystals of iridium ( @xmath45 ) , boron ( rhomboedral ) and magnesium ( @xmath52 ) may be helpful for choosing sensible phonon frequency range for our discussion . at first we assume , that the same type of atoms , occupying different sites in the cell , have similar average vibrational frequencies .
iridium , as the heaviest element , is expected to have the lowest @xmath19 , debye frequency of metallic iridium is about @xmath53 mev ( 420 k ) @xcite . in contrast
, the lightest boron is certainly expected to have the highest phonon frequencies , and for crystalline boron @xmath54 mev ( 1200 k ) @xcite .
finally , magnesium average frequencies are expected to locate between the values of iridium and boron .
debye frequency of crystalline mg is rather low : @xmath55 mev ( 400 k ) @xcite .
we may also recall , that average square phonon frequency is often estimated as @xmath56 , which is a good approximation in monoatomic structures . in our case , these values may also be helpful for choosing feasible range of @xmath19 .
presents estimated electron - phonon coupling constant @xmath3 , associated with particular crystal sites , plotted as a function of average square phonon frequency . for each type of atoms , a wide frequency range was chosen , to illustrate the changeability of @xmath3 . in the case of iridium ,
the largest contribution to @xmath3 comes from ir3 sublattice , due to the large population of this site . among boron and magnesium sites ,
the b2 and mg1 atoms provide the largest contributions .
however , one has to remember that this comparison is valid only if we assume identical frequencies for the same atoms at different sites . in this simplified analysis
, we may also plot the overall coupling constant for constituent atoms , by adding the contributions from each site , which is illustrated by solid lines in .
in mg@xmath0ir@xmath1b@xmath2 from iridium ( a ) , boron ( b ) and magnesium ( c ) as a function of a root of average square frequency .
top curve on each graph is a sum of contributions from particular sites .
vertical lines mark the value of @xmath57 . ] as we can see , the obtained partial coupling constants are not high for each atom . at the moment , we are not aware of any experimental findings of the epc constant in this compound
. however , we can try to estimate the range of `` experimental '' @xmath3 , analysing the magnitude of the observed critical temperature .
if we assume , that we are dealing with bcs - type superconductivity , we may substitute the experimental value of @xmath58 k into the mcmillan formula for @xmath59 @xcite : @xmath60.\ ] ] however , because the debye temperature @xmath51 of mg@xmath0ir@xmath1b@xmath2 is not known yet , we plot @xmath3 in as a function of @xmath51 , for typical values of coulomb pseudopotential parameter @xmath61 . the resulting epc constant @xmath3 varies between 0.5 and 0.75 , and for e.g. @xmath62 k we obtain @xmath63 , whereas for @xmath64 k we get @xmath65 ( for @xmath66 ) , as one can see in the .
thus , if mg@xmath0ir@xmath1b@xmath2 is treated as a conventional superconductor , the lower limit of @xmath3 is expected to be 0.5 . in order to get this value from our rmta calculations ,
relatively low average frequencies for all constituent atoms are required . if we take the following @xmath67 : , and , we obtain @xmath68 , @xmath69 and @xmath70 , which gives the expected lower limit of epc constant ( @xmath71 ) .
if higher values of @xmath3 are experimentally observed , it will indicate either additional contributions , neglected within the rmta framework , or even lower phonon frequencies , than the values used in the aforementioned estimations .
especially iridium contribution is sensitive to the change of @xmath72 parameter , i.e. increasing it to 25 mev ( probably an upper limit ) results in decrease of @xmath73 to 0.1 , that is to the value found for other sublattices .
generally , our calculations indicates , that iridium sublattices seem to be the most important for the onset of superconductivity in mg@xmath0ir@xmath1b@xmath2 . ,
evaluated from the mcmillan formula , using observed @xmath58 k , plotted as a function of debye temperature @xmath51 for three values of coulomb pseudopotential parameters @xmath61.,scaledwidth=50.0% ] finally , we shortly comment the possible influence of the lack of inversion symmetry and spin - orbit ( so ) interaction on the electronic structure and superconductivity .
the absence of a centre of inversion in a superconductor affects the symmetry of superconducting state , allowing for an admixture of singlet and triplet components @xcite .
because the singlet pairing is based on the time - reversal symmetry @xcite , which is present as long as the compound is not magnetic , mainly the triplet channel is affected by the lack of inversion , and the superconductivity may be even suppressed , see e.g. @xcite .
the so interaction , which mixes the initial spin - up and spin - down electronic states , plays an additional role .
it was found , that it may control the mixing of parity of the superconducting state @xcite .
this seems to be the case of li@xmath5pd@xmath4b and li@xmath5pt@xmath4b compounds , where specific heat @xcite and nmr @xcite measurements strongly support phonon - mediated isotropic superconductivity , while penetration depth measurements suggests an admixture of spin - singlet and triplet components in the superconducting energy gap @xcite , with larger triplet component in the pt - case .
this kind of experimental study for mg@xmath0ir@xmath1b@xmath2 should be prior to the theoretical discussion of gap symmetry in this compound .
as far as the band structure of mg@xmath0ir@xmath1b@xmath2 is concerned , the modifications due to the so interaction are not expected to significantly affect the obtained values of dos and mh parameters . for metallic @xmath45 iridium
, our value of mh parameter , calculated also neglecting so interaction , @xmath74 mry/(a.u.)@xmath75 , gives the correct magnitude of the electron - phonon coupling constant @xmath76 , comparing to the observed @xmath77 @xcite .
here , the average square frequency was estimated from the formula @xmath56 , using the debye temperature @xmath78 k.
the results of lda electronic structure calculations of new mg@xmath0ir@xmath1b@xmath2 superconductor were presented .
the main contributions to densities of states near @xmath23 are provided by iridium atoms .
the electron - phonon coupling constant @xmath3 was roughly estimated , using the calculated mcmillan - hopfield parameters and qualitative discussion of average phonon frequencies .
we discussed the relation of the experimental transition temperature and the magnitude of epc coupling . within the rigid - muffin - tin approximation ,
the main contribution to @xmath3 comes from iridium , with smaller contributions from boron and magnesium .
if any information about the dynamic properties of atoms in this compound become available , our analysis presented of will allow to find better theoretical estimation of @xmath3 .
the location of @xmath23 on the slope of ir dos peak leads to rough suggestion , that hole doping on iridium sites , e.g. with rhodium , may increase the densities and mh parameters .
we would like to thank dr tomasz klimczuk for helpful discussions .
this work was partly supported by the polish ministry of science and higher education ( phd grant ) .
30 bauer e , hilscher g , michor h , paul ch , scheidt e w , gribanov a , seropegin yu , nol h , sigrist m and rogl p 2004 027003 akazawa t , hidaka h , fujiwara t , kobayashi t c , yamamoto e , haga y , settai r and onuki y 2004 l29 togano k , badica p , nakamori y , orimo s , takeya h and hirata k 2004 247004 badica p , kondo t and togano k 2005 1014 anderson p w 1984 4000 frigeri p a , agterberg d f , koga a and sigrist m 2004 097001 samokhin k v , zijlstra e s and bose s k 2004 b * 69 * 094514 klimczuk t , xu q , morosan e , thompson j d , zandbergen h w and cava r j 2006 b * 74 * 220502(r ) lee k w and pickett w e 2005 b * 72 * 174505 bansil a , kaprzyk s , mijnarends p e and tobola j 1999 b * 60 * 13396 stopa t , kaprzyk s and tobola j 2004 4921 von barth u and hedin l 1972 1629
mcmillan w l 1968 331 hopfield j j 1969 443 gaspari g d and gyrffy
b l 1972 801 gomersall i r and gyrffy b
l 1974 1204 klein b m and papaconstantopoulus d a 1974 1193 pickett w e 1982 b * 25 * , 745 wiendlocha b , tobola j and kaprzyk s 2006 b * 73 * 134522 wiendlocha b , tobola j , kaprzyk s , fruchart d and marcus j 2006 b * 243 * 351 kittel c 1996 _ introduction to solid state physics _ ( new york : john wiley & sons ) thompson j c and mcdonald w j 1963 82 anderson p w 1959 26 takeya h , hirata k , yamaura k , togano k , el massalami m , rapp r , chaves f a and ouladdiaf b 2005 b * 72 * 104506 nishiyama m , inada y and zheng g 2005 b * 71 * 220505(r ) yuan h q , agterberg d f , hayashi n , badica p , vandervelde d , togano k , sigrist m and salamon m b 2006 017006 a. kokalj 1999 j. mol
. graphics modelling * 17 * , 176 .
code available from http://www.xcrysden.org/. | electronic structure of a novel superconducting noncentrosymmetric compound mg@xmath0ir@xmath1b@xmath2 was calculated using the korringa - kohn - rostoker method .
electronic part of the electron - phonon coupling constant , mcmillan - hopfield parameters , were calculated using the rigid - muffin - tin approximation ( rmta ) .
the magnitude of the electron - phonon coupling constant @xmath3 , analysing atomic contributions , is discussed .
our results show , that superconductivity in mg@xmath0ir@xmath1b@xmath2 is presumably mediated by electron - phonon interaction . |
this is a final paper of the series ( @xcite ; dogiel et al .
2009abc ) on energetic processes in the galactic center ( gc ) .
in these papers we presented our interpretation of x - ray and gamma - ray emission from the galactic center .
we supposed that all these phenomena had common origin and were initiated by accretion processes onto the central supermassive black hole .
we showed that the the observed x - ray continuum and line emission from the gc might be produced by a flux of subrelativistic protons which resulted from an unbounded part of stars accreted onto the central black hole .
we estimated the average energy of escaping protons to be about 100 mev in order to produce a flux of hard x - ray emission in the energy range @xmath0 kev as observed by suzaku / hxd @xcite .
the quasi - stationary production rate of subrelativistic protons hardly exceeds @xmath4 protons s@xmath3 for the frequency of star capture @xmath5 years@xmath3 @xcite and a fraction of escaped matter equaled @xmath6% of star masses @xcite .
the unbounded fraction of stars escapes with subrelativistic velocities which correspond energies about 100 mev for protons and 50 kev for electrons ( see @xcite ) .
the numbers of protons and electrons equal to each other .
therefore , fluxes of bremsstrahlung hard x - ray emission produced by protons and by electrons equal to each other too . but
the lifetime of electrons with energies 50 kev is in about five orders of magnitude smaller than that of 100 mev protons and the electron bremsstrahlung flux is significant for a short time just after a capture event .
since it is supposed that relatively long time has passed after the last star capture , electron bremsstrahlung radiation is negligible .
our goal is to estimate the spatial diffusion coefficient , @xmath7 , of subrelativistic particles near the gc , whose value is unknown . below we derive its value from the spatial distribution of hard x - ray emission near the gc as observed by suzaku . in section 2 ,
we summarize an energetics of the galactic center diffuse emission . in section 3 , we carry forward our model using non - uniform target gas distribution .
section 4 is devoted to explain how the model and the hxd data can be compared . in section 5 and 6 , spatial distribution of 6.7
kev fe line emission and hard x - ray continuum are compared with ones expected from the present model , deriving a confinement on a diffusion coefficient for the sub - relativistic protons .
@xcite reported that the hard x - ray spectrum in the range @xmath8 kev from the gc is naturally explained by a 6.5 kev - temperature plasma plus a power - law component with the photon index of @xmath9 .
latter on , @xcite found a prominent hard x - ray emission in the range from 14 to 40 kev whose spectrum is a power law with the spectral index ranging from 1.8 to 2.5 .
different processes may contribute to the total emission from the gc .
_ thermal emission_. the emissivity of thermal bremsstrahlung of a hot plasma @xmath10 can be estimated from the equation ( see e.g. @xcite ) @xmath11 where the temperature is in kelvin degrees .
as follows from the analysis of suzaku data the plasma temperature in the gc is constant and equals 6.5 kev @xcite .
then the thermal emissivity in the gc is @xmath12 @xcite presented results of calculation for the @xmath13 central region and showed that for the plasma density @xmath14 @xmath15 and the temperature 6.5 kev the total thermal flux from this region in the energy rage @xmath0 kev was about @xmath16 erg s@xmath3 .
the total non - thermal flux was estimated by the value @xmath17 erg s@xmath3 .
_ non - thermal component_. @xcite found that a combination of the fe xxv - k@xmath18 and fe i - k@xmath18 fluxes was proportional to the non - thermal continuum flux in the 5 - 10 kev energy band .
they concluded that the total non - thermal flux consists of the two components , one of which is proportional to the intensity of 6.7 kev line , and , therefore , is produced in regions of hot plasma , while the intensity of the second component is proportional to that of the 6.4 kev line and is generated in regions of molecular gas .
@xcite reported a similar conclusion obtained from the xmm - newton observations . in the model of x - ray production by subrelativistic protons this relation
is naturally explained ( see dogiel et al .
continuum and line emission is produced in this case by interactions of subrelativistic protons with the background plasma and the molecular hydrogen in the gc . in spite of relatively small radius ( @xmath19 pc )
the inner galactic region contains about 10% of the galaxy s molecular mass , @xmath20 .
most of the molecular gas is contained in very compact clouds ( see the review of @xcite ) .
since the total mass of the molecular gas is much larger than that of the hot plasma , @xmath21 , one may assume that most of inverse bremsstrahlung flux produced by protons would come from regions filled by the molecular gas .
however , as @xcite showed , the mean free path of protons in dense molecular clouds is very short since the diffusion coefficient inside molecular clouds is quite small .
therefore , only a small fraction of the molecular hydrogen is involved into processes of inverse bremsstrahlung radiation .
thus , for the molecular cloud sgr b2 with the mass about @xmath22 it follows from calculations of @xcite and the suzaku data ( see @xcite ) that the total bremsstrahlung flux in the range @xmath8 kev is about @xmath23 erg s@xmath3 .
the total flux of the 6.4 kev line emission from sgr b2 is about @xmath24 erg s@xmath3 that gives the ratio @xmath25 . for higher energies of x - ray photons
our calculations show that the ratio of @xmath26 . the total flux of the 6.4 kev line from the gc can be estimated from the asca data @xcite and for the region @xmath27 it is about @xmath28 erg s@xmath3
, then we expect that the total non - thermal bremsstrahlung flux from proton interaction with the molecular gas is about @xmath29 erg s@xmath3 , and 70% of the total non - thermal bremsstrahlung emission in the range @xmath0 kev , @xmath30 erg s@xmath3 , is generated in regions filled with the hot plasma .
in this section we present a qualitative effect of plasma density variations on the proton distribution in the gc .
we remind that the spatial distribution of subrelativistic protons is described by the quasi - stationary diffusion equation @xmath31 where @xmath32 is the density of cosmic rays , @xmath7 is the spatial diffusion coefficient , @xmath33 is the rate of energy losses which for subrelativistic protons is determined by the coulomb collisions and has the form @xmath34 where @xmath35 is the proton velocity , @xmath36 is the coulomb logarithm , and @xmath37 is the plasma density , which is a function of the radius @xmath38 in general case .
the source function is taken in the form @xmath39 where @xmath40 protons s@xmath3 and @xmath41 mev is the injection energy of protons . for @xmath42 ,
the spherically symmetric solution is @xmath43 \label{n_uni}\ ] ] where @xmath44 in the case when the plasma density is spatially variable as @xmath45 where @xmath46 is a scale parameter , the analytical solution for @xmath47 is @xmath48 \exp\left[-\frac{{\ln^2\left(r / r_0\right)r_0 ^ 2}}{4\lambda(e)}\right ] \label{n_nuni}\ ] ] here @xmath49 in figure [ model_distr ] we show the difference between uniform for @xmath50 @xmath15(solid line ) and non - uniform cases for the parameters @xmath51 @xmath15 ( dashed - dotted line ) and @xmath52 @xmath15 ( dashed line ) . the scale parameter is taken to be @xmath53 pc .
one can see that variations of the gas density change significantly the proton distribution .
in section 5 and 6 , we will compare spectra expected from the present model with the hxd spectrum observed around the gc ( @xcite ) . before describing the comparison result , we present methods and assumptions which were used in the calculation .
the intensity @xmath54 of inverse bremsstrahlung emission of protons in any direction @xmath55 is calculated from @xmath56 where the integration is along the line of sight @xmath55 and @xmath57 is the energy of photons . here
@xmath58 is the density of subrelativistic protons with the energy @xmath59 at the radial distance @xmath38 from the gc , @xmath60 is the velocity of protons with the energy @xmath59 , @xmath61 is the cross - section of inverse bremsstrahlung radiation , and @xmath62 is the plasma density distribution in the gc . to compare such results of our model calculations , which are given in _
photons @xmath63kev@xmath3s@xmath3sr@xmath3 _ with the suzaku data of @xcite given in _
counts s@xmath3 _ , we should convolve the hxd energy and angular responses ( see @xcite ) to the model intensity .
the procedure is as follows : @xmath64 where @xmath65 is in ( _ counts _ _ s_@xmath3 ) denotes the expected count rate in the direction determined by the galactic coordinates @xmath66 , @xmath67 is in ( _ sr _ ) , @xmath68 in ( _ ph @xmath63kev@xmath3s@xmath3sr@xmath3 _ ) is the continuum intensity in the direction @xmath69 , @xmath70 and @xmath71 represent the effective area and angular transmission of the hxd , respectively .
the hxd data can still include a contaminating flux from a number of x - ray point sources besides the diffuse emission as noted in dogiel et al .
( 2009b ) .
the flux level of the contamination is calculated to be on the order of 10% by integrating the known @xmath72 curve for x - ray point sources around the gc obtained with chandra @xcite over the luminosity of of @xmath73 erg s@xmath3 in which range has been actually measured so far ( in the `` field '' region of @xcite ) .
although the contribution can be larger if the @xmath72 curve is measured further smaller luminosities , in the present comparison with the model calculation , we simply ignore their contribution .
the effect of this neglection is discussed later in section 6 .
the spatial distribution of the hot plasma in the gc can be derived from the distribution of the 6.7 kev iron line which traces the hot plasma .
the origin of the k-@xmath18 iron line from the hot plasma of the gc is pure thermal ( see @xcite ) .
then for the surface brightness distribution of the 6.7 kev emission line in any direction @xmath55 to the gc we have @xmath74 @xcite obtained a surface brightness distribution of the 6.7 kev emission line , @xmath75 , from the asca data as a function of the angle from the galactic center , @xmath76 , @xmath77 where @xmath78 is the angle from @xmath79 , @xmath80=19.7 and @xmath81=1.6 in _ photons @xmath63s@xmath3sr@xmath3 _ units , and @xmath82 and @xmath83 in degrees .
this distribution is shown in figure [ 6.7 ] by the thick solid line .
( 110mm,80mm)6.7_asca.eps from the 6.7 line distribution of @xcite we try to derive the plasma density distribution in the gc .
then we will use this density distribution for calculations of hard x - ray emission from the gc .
we represent the hot plasma distribution by analytical functions as @xmath84\ , , \label{nr}\ ] ] and our goal is to find suitable parameters @xmath46 and @xmath18 which give good correspondence of the observed asca data and that obtained with the density distribution ( [ nr ] ) .
( 110mm,80mm)fig_1.eps in order to derive the model spatial variations of the 6.7 kev line emission along the galactic plane we should integrate along the direction @xmath55 shown in figure [ fig_1 ] .
the angle between the direction to the gc and the line of view is the galactic longitude @xmath85 .
the circle line around the galactic center defines a sphere filled with a hot plasma whose radius is supposed to be about @xmath86 pc in accordance with the x - ray observations ( see , e.g. @xcite ) .
the radius - vector @xmath38 from the gc along the line of sight as a function of the galactic longitude @xmath87 and the angle @xmath88 between the line coinciding earth an the gc ( arrow line ) is described as @xmath89 then with the density distribution ( [ nr ] ) the integration along the line of sight in the direction * s * is @xmath90 where @xmath91 and the angles @xmath92 shown by the two radius - vectors in figure [ fig_1 ] is @xmath93\ ] ] here @xmath94 kpc is the distance between earth and the gc which equals 8 kpc @xcite . from equation ( [ 6.7_mod ] ) and the asca 6.7 kev spatial variations obtained by @xcite we derived parameters @xmath46 and @xmath18 , that gives @xmath95 pc and @xmath96 .
the asca spatial variations of 6.7 kev line are shown in figure [ 6.7 ] by the thick solid line .
as an example we show also in this figure the model spatial variations of the 6.7 kev line intensity derived for @xmath53 pc and @xmath97 ( dashed - dotted line ) .
the coincidence between calculations and data are reasonably good . for comparison we showed also in this figure the expected 6.7 kev line spatial variation for the case of uniform plasma distribution ( thin solid line ) .
one can see that the asca data can hardly be described by a model with the uniform plasma distribution .
the parameter @xmath98 of equation ( [ nr ] ) can be roughly estimated by comparing the model emission ( 6.5 kev thermal + non - thermal ) with the observed spectrum of the hard x - ray emission around the gc ( @xcite ) . to perform such a comparison , we convolved the model spectrum with the hxd response as described in section 4 , and then , overlaid it over the hxd spectrum by adjusting the model normalization factor , @xmath98 in equation ( [ nr ] ) , so that the model reproduces the hxd fluxes .
figure [ xr - sp](a ) shows a result of the convolution , and the model well reproduce the observed spectrum in the @xmath0 kev band .
although we only present the result for one parameter set explained in the caption of figure 4 , the model spectral shape does not strongly depend on the parameters , and within possible parameter ranges , @xmath98 was calculated to be @xmath99 @xmath15 .
( 80mm,80mm)count_spec.eps ( 80mm,80mm)ph_spec.eps as one can see from these figures the contributions of thermal and non - thermal emission into the total flux from the gc are almost equal to each other in this energy range ( their ratio is 2:3 ) .
in order to calculate the spatial distribution of subrelativistic protons in the gc we should take into account energy losses in regions of hot plasma and in the regions of neutral gas surrounding the gc whose average density was taken to be @xmath100 @xmath15 .
as follows from the analysis of @xcite the spatial distribution of the intensities of thermal and non - thermal emission in the gc are similar to each other .
since the first one is proportional @xmath101 and the second one is proportional to @xmath102 , where @xmath103 is the spatial distribution of subrelativistic protons , it means that the spatial distribution of subrelativistic protons does not differ strongly from that of plasma . the expected spatial distributions of the @xmath0 kev x - ray emission in the gc derived from equation ( [ is ] ) for @xmath104@xmath2s@xmath3 ( solid line ) and for @xmath105@xmath2s@xmath3 ( dashed line ) and the suzaku data obtained by @xcite are shown in figure [ f3 ] . , the data point at @xmath106 is largely discrepant from the model , and even from the other data points . as noted in @xcite
, this deviation could be due to underestimation of contaminating signals from known bright point source inside the hxd / pin fov .
therefore , this point should be ignored in the present comparison between observed data and the model calculation . ]
our calculations show that spatial distribution of the non - thermal @xmath0 kev x - ray emission does not differ significantly from each other .
however , the necessary proton production rate @xmath107 is an essential function of the diffusion coefficient .
thus , for @xmath104@xmath2s@xmath3 the value of @xmath107 needed to reproduce the suzaku data equals @xmath108protons s@xmath3 while for @xmath105@xmath2s@xmath3 the production rate is necessary to be @xmath109protons s@xmath3 .
we remind that the average rate of proton production by accretion on the black hole is no more than @xmath110protons s@xmath3 .
since the value of q is restricted , then permitted values of @xmath7 are confined within a relatively narrow range around @xmath111 @xmath2 s@xmath3 .
however , these estimates were obtained in assumption that almost 100% of the @xmath0 kev x - ray flux from the gc is generated by subrelativistic protons .
as follows from , e.g. , @xcite a significant part of the gc hard x - ray flux may be due to faint point sources .
then a smaller production rate of protons in the gc is required , and , hence , higher values of the diffusion coefficient can not be excluded .
analysis of x - ray data @xcite demonstrated that continuum non - thermal emission in the gc consisted of two components one of which was proportional to the intensity of 6.7 kev iron line which traced distribution of hot 6.5 kev plasma , and the other one was proportional to the intensity of 6.4 kev plasma which traced distribution of cold molecular hydrogen .
if this non - thermal emission is due to inverse bremsstrahlung emission of subrelativistic protons , then their interactions with hot and cold fractions of the interstellar medium are equally important .
our estimation show that about 30% of the total non - thermal flux from the gc in the range @xmath0 kev is generated in regions of cold gas while the rest is produced by proton interaction with hot plasma .
the reason is that only a small fraction of molecular gas is involved into processes of bremsstrahlung emission because of short mean free path of protons in molecular clouds . from the x - ray data obtained with asca
@xcite it was concluded that the plasma density distribution in the gc is strongly non - uniform , e.g. the model distribution of 6.7 kev line in the gc obtained with the plasma profile @xmath112(\mbox{cm$^{-3}$})\ ] ] is very close to that from asca observation .
this effect of density variations should be included into analysis of protons propagation in the gc . from the total spectrum of x - ray emission in the range @xmath0 kev from the @xmath113-diameter central region and the spatial variation of this emission within @xmath114-diameter central region
as observed by suzaku and asca we derived the characteristic diffusion coefficient of subrelativistic protons .
it was shown that the diffusion coefficient of subrelativistic particles is about the value of @xmath115 @xmath2s@xmath3 . for higher values of the diffusion coefficient a higher production rate of subrelativistic protons
is necessary which can not be provided by accretion .
however , this estimation of @xmath7 one should consider as a lower limit .
the point is that we assumed here that all hard x - ray emission in the @xmath0 kev band is produced by subrelativistic protons in the hot fraction of the interstellar gas .
however as stated above a part of this emission may be produced in molecular clouds .
besides , we do not know exactly which part of this emission is due to unresolved point sources ( see discussion in @xcite ) . therefore , we can not exclude that the diffusion coefficient in the gc is close to its average value in the galaxy , i.e. @xmath116 @xmath2 s@xmath3 .
thus , we determine the range of values of the diffusion coefficient in the gc as @xmath117@xmath2s@xmath3 .
vad and doc were partly supported by the rfbr grant 08 - 02 - 00170-a , the nsc - rfbr joint research project no 95wfa0700088 and by the grant of a president of the russian federation `` scientific school of academician v.l.ginzburg '' .
ksc is supported by a rgc grant of hong kong government under hku 7014/07p .
a. bamba is supported by jsps research fellowship for young scientists ( 19 - 1804 ) .
cmk is supported in part by national science council , taiwan under the grant nsc-96 - 2112-m-008 - 014-my3 .
dogiel , v.a . , cheng , k .- s . ,
chernyshov , d.o .
2008 , new astronomy reviews , 52 , 460 dogiel , v. a. , tatischeff , v . , cheng , k .- s . , et al .
2009a , a&a , submitted dogiel , v. , chernyshov d. , yuasa , t. et al .
2009b , pasj , submitted dogiel , v. a. , cheng , k - s . , chernyshov , d.o .
et al . 2009c , pasj , to be published in pasj , 61 , no.5 donley , j. l. , brandt , w. n. , eracleous , m. , & boller , th .
2002 , apj , 124 , 1308 .
ginzburg , v.l .
1989 , _ applications of electrodynamics in theoretical physics and astrophysics _ , gordon and brech science publication .
kokubun , m. , makishima , k. , takahashi , t. et al .
2007 , pasj , 59 , 53 koyama , k. , maeda , y. , sonobe , t. et al .
1996 , , 48 , 249 koyama , k. , inui , t. , hyodo , y. et al . 2007a , pasj , 59 , 221 koyama , k. , hyodo , y. , inui , t. et al .
2007b , pasj , 59 , s245 reid , m. j. 1993 , ara&a , 31 , 345 revnivtsev , m. , sazonov , s. , churazov , e. et al .
2009 nature , 458 , 1142 syer , d. , & ulmer , a. 1999 mnras , 306 , 35 takahashi , t. , abe , k. , endo , m. et al .
2007 , pasj , 59 , 35 | we showed that if the non - thermal emission from the galactic center in the range @xmath0 kev is due to inverse bremsstrahlung emission of subrelativistic protons , their interactions with hot and cold fractions of the interstellar medium are equally important .
our estimation show that about 30% of the total non - thermal flux from the gc in the range @xmath0 kev is generated in regions of cold gas while the rest is produced by proton interaction with hot plasma . from the spatial distribution of 6.7 kev iron line we concluded the spatial distribution of hot plasma is strongly non - uniform that should be taken into account in analysis of protons propagation in the gc . from the suzaku data we got independent estimates for the diffusion coefficient of subrelativistic protons in the gc , which was in the range
@xmath1 @xmath2s@xmath3 |
the galactic center region is known to have a high concentration of massive , warm , dense and turbulent molecular clouds . due to strong tidal forces exerted by the gravitational potential of the nuclear cluster and the massive black hole , sgr a * ,
only dense molecular gas is expected to survive in this region ( see * ? ? ?
this has the ramification that the formation of massive stellar clusters must be pervasive in this region as evidenced by the arches and quintuple clusters as well as the nuclear massive cluster centered on sgr a * ( * ? ? ?
* ; * ? ? ?
* and the references therein ) .
there are also young clusters of o - b stars traced by ultracompact ( uc ) hii regions in sgr b2 ( e.g. * ? ? ?
* ) which is possibly the best example of mini - starburst activity in the galaxy .
although our present understanding of star formation is primarily limited to observations of low - mass stars @xcite , probing star forming sites in the galactic center region can be useful in addressing the key mechanism by which massive stars are formed ( m @xmath8 10 ) ( see * ? ? ? * and references cited therein ) . here
, we examine the nature of a group of massive young stars lying within 2.5@xmath9 of the dynamical center of the galaxy @xcite . at a projected distance of 6 pc from sgr a * ,
there is a cluster of compact hii regions on the edge of the 50 molecular cloud m-0.02 - 0.07 .
this cluster consists of brightest hii regions in the galactic center with the exception of the cluster of hii regions in sgr b2 .
the 50 molecular cloud itself is interacting with the nonthermal sgr a east ( e.g. * ? ? ?
* and references cited therein ) which is known to be a supernova remnant ( snr g0.0 + 0.0 ) . since sgr a east appears to be interacting with the galactic center circumnuclear molecular ring , which itself orbits sgr a * , it is probable that sgr a east , m-0.02 - 0.07 , and the cluster of hii regions all lie near the galactic center , not far from their projected distances .
the compact hii regions , known as the sgr a east hii a to d ( sgr a a - d ) , trace the site of recent massive star formation nearest to the dynamical center of the galaxy .
the four hii regions have radial velocities ranging between 43 and 49 , thus appear dynamically coupled to the 50 molecular cloud @xcite . with the exception of [ ariii ] emission from sgr a - d hii source , ir spectroscopic measurements detected [ neii ] and [ ariii ] line emission from all four components ( a - d ) indicating that the exciting stars have spectral types of o8-b0 @xcite , consistent with h76@xmath2 measurements @xcite .
the measured extinction toward sgr a east a - c is less than that of source d @xcite .
additional extinction for source d is likely due to the 50 molecular clouds or due to material associated with the ionizing stellar source @xcite .
these authors estimate the extinction values at 12.8@xmath1 m are @xmath10mag for a - c and @xmath03.3 toward the sgr a d source .
these extinction measurements imply that the sgr a east a - c hii regions , which are resolved spatially , are located on the front side of the molecular cloud , whereas the compact sgr a d source or g359.96 - 0.08 ( hereafter this source is called d ) , may be embedded within the 50 km s@xmath7 molecular cloud .
near - ir measurements have also examined the stellar sources associated with these hii regions @xcite . a highly reddened source with a h - k@xmath11 @xmath0 3.5
is detected at the position of the most compact hii source d. the low fractional abundance of ar@xmath12 led to the suggestion that stellar temperatures of the sgr a east hii regions range between 34,500 and 36,500k @xcite which are consistent with spectra types o8-o9 stars @xcite .
the spectral identification of the sgr a d source has not been clear but @xcite suggest that its spectrum could be consistent with a combination of an ultracompact ( uc ) hii region and b[e ] stars . here
, we investigate radio and infrared properties of the most compact ( d source ) and extended ( a source ) of the sgr a east hii regions .
radio continuum observations were made with the d configuration of the very large array ( vla ) of the national radio astronomy observatory ( nrao ) , at 43 ghz ( 7 mm ) .
the source was observed as part of a survey of uc hii regions in the galactic center , which took place may 29 , 2009 .
the source was observed in the fast switching mode of the vla ( each cycle had 150 seconds on the source and 15 seconds on the complex gain calibrator - nrao530 ) for a total of 20 minutes .
3c286 was used as a flux calibrator .
standard data editing and calibration of the gains was done in aips . using standard calibration
, we also reduced multi - configuration radio continuum data at 8.3 ghz ( ay43 ) .
the 4.8ghz data were taken from @xcite . the sgr a east hii region cluster was observed in the [ ne ii ] ( 12.8 @xmath1 m ) line with the high - resolution mid - infrared spectrograph texes on the nasa irtf on 2009 june 3 and 2010 may 30 ( ut ) .
texes @xcite has several operating modes .
these observations were made in two modes : high - resolution , cross - dispersed , with velocity resolution @xmath04 and spatial resolution @xmath01.2 along a [email protected] slit , and medium resolution , long - slit , with velocity resolution @xmath025 along a 1.4@xmath1355 slit .
the pixel spacing is 0.36 spatially and 0.95 spectrally in cross - dispersed mode .
it is 0.36 and 8.9 in long - slit mode .
maps were made by stepping the n - s oriented slit to the east , with 0.7 steps in 2009 and 0.35 steps in 2010 . the sgr a east hii region was mapped in 2009 in cross - dispersed mode by making a series of scans , each covering [email protected] ( ra@xmath14dec ) , which were pieced together to cover a 34@xmath1368 region ( with some small gaps ) , with the help of a long - slit scan that covered the entire region .
hii source d was reobserved in 2010 to obtain better spatial sampling and higher signal - to - noise ratio .
these measurements also allowed us to detect 12.8@xmath1 m emission from the d source using our low spectral resolution data and medium resolution data taken in 2009 .
the spectral coverage of the data was 2270 , which in the cross - dispersed observations was broken into 8 spectral orders of the high - resolution grating . [ ne ii ] line emission was detected in a 60 interval centered at v@xmath15 @xmath16 46 .
dust continuum emission was only marginally detected outside of this spectral interval ( also see figure 3 showing source d before and after deconvolution . ) the data were first processed with the standard texes pipeline reduction program @xcite , which removes spikes , corrects for optical distortions , and produces wavelength and intensity calibrated maps .
it was then processed with a maximum - entropy - method deconvolution routine , which improved the spatial resolution to about 0.7 , slightly better than the infrared telescope facility ( irtf ) diffraction limit . because a part of the point - spread function is caused by seeing , which can not be exactly the same for the observations of the psf calibrator ( vx sgr ) and the hii region , the results of the deconvolution should be viewed with some caution , but the close similarity between the deconvolved [ ne ii ] data and the higher resolution radio continuum data indicates that the results should be reliable .
in addition , all of the features that we discuss in the [ ne ii ] data can be seen before deconvolution , although some sources , notably the two peaks in hii source d , blend together .
the absolute astrometry of the 12.8@xmath1 m data is about few arcseconds .
we obtained coordinates for the [ ne ii ] maps by aligning them with the radio maps
. these should be good to better than 0.35 " ( 1 pixel ) .
the relative astrometry of the [ ne ii ] and 12.8um continuum should be better since they were observed simultaneously .
pa@xmath2 emission - line and continuum images of the sgr a east region were obtained as part of the larger survey of the galactic center by @xcite , using the near - infrared camera and multi - object spectrometer ( nicmos ) on hst ( program 11120 ) .
exposures were obtained in march may 2008 using the nicmos camera 3 , which has a pixel scale of 0.20 .
images were obtained in narrow - band filters centered on the pa@xmath2 emission line at 1.87@xmath1 m and the nearby continuum at 1.90@xmath1 m .
a pattern of dithered observations was obtained at each pointing in the survey with each exposure having a duration of 48 seconds . for this study we used a total of 8 exposures in each of the two filters that covered the region near source d. calibrated images
were retrieved from the hubble data archive , which had been processed using the latest nicmos instrumental calibrations and data reduction software .
the eight calibrated images in each filter were combined using the `` multidrizzle '' task in the stsdas package of pyraf .
the final drizzle - combined images at pa@xmath2 and 1.90@xmath1 m were subsequently rescaled using the latest nicmos photometric zeropoint information to produce images in calibrated units of jy . to determine the position uncertainty of the stars ,
a gaussian was fitted to two different f190n images with different exposures .
both images were used in making the drizzled mosaic .
this allows us to to measure random pointing errors from exposure to exposure , as well as the uncertainties in our measurements of the centroid in the images .
the random errors in pointing and measurements are on the order of @xmath17 .
this is entirely consistent with the random relative errors associated with hst reacquiring guide stars in between exposures .
in addition to this random error of @xmath18 , there is also an absolute rms uncertainty of @xmath19 in how well the coordinates of the hst guide stars are known .
thus , the coordinates of the hst images can be trusted only at the level of 0.2 .
irac observed the sgr a d as part of the galactic center ( galcen ) survey by @xcite .
it was covered by 5 exposures with 1.2 sec frame times .
the irac flux densities for sgr a d were obtained from both the galcen catalog @xcite and the independently reduced glimpse ii catalog ( an extension of the glimpse survey , @xcite ) . at 3.6 , 4.5 , and 5.8 @xmath1 m we use the glimpse ii measurements , which tend to be @xmath20 lower than the galcen measurements .
the galcen catalog indicates that the source is near or at saturation at all wavelengths , although at 3.6 @xmath1 m the source is only slightly past the nominal saturation limit .
the glimpse ii catalog does not include an 8 @xmath1 m flux density . at this wavelength ( and only this wavelength )
the irac images show a local minimum at the peak of the source which is characteristic of a strongly saturated source . we have attempted to apply t. jarret s iracworks software to recover the flux of sgr a d from the saturated images . at 3.6 - 5.8
@xmath1 m iracworks does not recognize sgr a d as a saturated source , suggesting that cataloged flux densities are not less than @xmath21 of the true values . at 8 @xmath1 m iracworks
successfully fits the source for only 2 out of 5 of the exposures , but does find a consistent result of @xmath22 mjy , which is about 2.5 times higher than the lower limit provided by the galcen catalog . reported magnitudes and flux densities are listed in table 1 .
llll 2mass 17455154 - 2900231 + & j & @xmath23 & @xmath24 + & h & @xmath25 & @xmath26 + & k & @xmath27 & @xmath28 + sstgc 0556837 ( galcen ) + & 3.6 & @xmath29 & @xmath30 + & 4.5 & @xmath31 & @xmath32 + & 5.8 & @xmath33 & @xmath34 + & 8 & @xmath35 & @xmath36 + sstglma g359.9672 - 00.0811 ( glimpse
ii ) + & 3.6 & @xmath37 & @xmath38 + & 4.5 & @xmath39 & @xmath40 + & 5.8 & @xmath41 & @xmath42 + & 8 & & + ( iracworks ) + & 8 & @xmath43 & @xmath22 + ( texes ) + & 12.8 & & 7500 @xmath44 [ tab : irac_2mass ]
the large - scale distribution of thermal and nonthermal emission from galactic center is complex .
however , within 4@xmath9 of sgr a * , the distribution of dust , molecular , ionized , and synchrotron emission shows that the 50 gmc m-0.02 - 0.07 interfaces with the supernova remnant sgr a east on one side and the cluster of hii regions on the opposite side .
figure 1a , b show the distributions of the 450 and 850@xmath1 m thermal emission from dust , superimposed on a 20 cm radio continuum image , respectively @xcite .
contours of submm emission arises from the eastern edge of the 50 molecular cloud .
these figures show that the cluster of compact hii sources sgr a a thru d are distributed in a region relatively devoid of emission at 450 and 850@xmath1 m .
the cavity in the submm distribution might be caused by a drop in dust temperature .
we made several cuts across the region where there is a depression of submm emission and determined a constant 450@xmath1 m to 850@xmath1 m flux ratio .
since the emission is optically thin , this ratio corresponds to a constant temperature . we note that dust temperature can not be constrained because it is entangled with the uncertainty in the ratio of optical depths . since there is no evidence for a temperature gradient in the dust distribution , it is possibile that the gas in the cavity is converted into formation of stars within the cloud .
source a is the most extended hii feature and lies to the north , whereas source d , the most compact hii source , lies to the south .
the distribution of the four hii regions appear to be anticorrelated with the distribution of dust and molecular gas , consistent with kinematic evidence that the 50 gmc is the parent cloud from which a cluster of massive young stars recently formed .
figure 2a shows a grayscale image of free - free continuum emission from the sgr a east hii cluster at 4.8ghz and figure 2b shows the integrated [ neii ] line intensity map . the spatial distribution of [ neii ] line emission matches very closely the free - free 4.8ghz continuum distribution , demonstrating that [ ne ii ] is a good tracer of the ionized gas .
figure 2c shows a contour map of source d before and after deconvolution demonstrating the elongated structure of source d in both maps .
figure 2d shows the line and continuum maps in two different panels .
the continuum map is from a 1200 interval blueward of the line .
the continuum map ( right panel of fig .
2d ) is narrower in the e - w direction than the line , and is centered on the peak of the line map within about 0.5@xmath45 of the centroid of the line emission . given the low resolution of the 2009 data , we can not determine the relative position of the continuum and line peaks with respect to each other .
the high resolution data can not detect the continuum emission . in this paper ,
we discuss two of the hii regions in the sgr a east group : d , the bright , compact source to the south , and a , the most extended source to the north .
sources b and c have morphologies and kinematics similar to source a , but are more irregular and fainter .
the brightest radio continuum source in figure 2 is source d , which is resolved for the first time into two equally bright radio continuum sources , d1 ( east ) and d2 ( west ) , separated by 12 at position angle ( p.a . )
@xmath16 70 .
each radio source consists of a compact source and a resolved structure elongated in the north - south ( ns ) direction .
contours of the emission at 4.87 ghz , as shown in figure 3a , extend for @xmath46 or @xmath00.7 pc , assuming that they are located at the galactic center distance of 8 kpc . the sgr a east hii regions have also been observed at 8.3 and 43 ghz .
figure 3b shows a grayscale image of source d at 43 ghz with a spatial resolution of 058 @xmath14 047 .
both sources d1 and d2 show n - s structure with an extent of 4- 5 .
the elongated features of sources d1 and d2 each show a compact component at 43 ghz .
table 2 give the integrated flux , the size and the peak flux of individual sources d1 and d2 at 4.87 , 8.3 and 43.3 ghz , respectively .
the gaussian fitted positions of d1 and d2 are @xmath47 and @xmath48 , respectively . the images from which the fitted data are obtained have identical spatial resolution 08 @xmath14 04 at all three frequencies .
the integrated fluxes s@xmath49 for source d1 and d2 given similar spectral index of @xmath50 ( where s@xmath51 ) between 8 and 44 ghz .
lrrrrrrrrrrrrrrrrrrrrrrrrr d1 & 4.87 & [email protected] & 1.7 & 1.2 & 6.6@xmath53 + `` & 8.45 & [email protected] & 1.7 & 0.7 & 8.6@xmath54 + '' & 44.34 & [email protected] & 1.5 & 1.0 & 8.7@xmath55 + + d2 & 4.87 & @xmath561.1 & 1.9 & 1.2 & 7.6@xmath53 + `` & 8.45 & [email protected] & 1.4 & 0.5 & [email protected] + '' & 44.34 & [email protected] & 1.1 & 0.5 & [email protected] + + from the radio images , d1 and d2 appear to be a pair of stars , each of which has a north - south bipolar jet .
this situation could occur if each star has a disk collimating its jet , and the two disks are coplanar , perhaps as a result of forming from the same cloud .
but for this to be the right model , the two stars would have to be nearly identical in their luminosities and jet properties , including a difference in their n and s jets , as well as having closely aligned disks .
this seems highly unlikely ( see additional arguments against two stars , as described below ) . a model with one exciting source for the entire ionized gas distribution would seem more plausible if it could explain the spatial distribution of radio emission .
to examine the nature of the exciting star or stars in source d further , we compared the radio images with a pa @xmath2 and 1.90@xmath1 m images of the region .
the cross drawn on figure 3a shows the peak position of the 1.90@xmath1 m emission .
the position of the stellar source does not change when both 1.90 and 1.87@xmath1 m images are compared with each other .
the centroids of the source at these wavelengths match to @xmath57 .
since the images in the these two filters were obtained back - to - back , without moving the telescope , the relative positioning should be within 10 - 20 milli - arcseconds .
the 1.90@xmath1 m emission appears to coincide close to the compact component of source d2 , suggesting that a single source in the vicinity of or coincident with d2 may be responsible for the ionization producing the extended radio continuum and pa@xmath2 nebular emission .
alternatively , it is possible that the extinction is variable locally being responsible for a lack of 1.90@xmath1 m detection from source d. in the continuum image the d2 source is very bright 5.1@xmath52 mjy while d1 is nearly undetected @xmath58 mjy . in the emission - line image ,
on the other hand , d1 and d2 sources have the flux of [email protected] and [email protected] mjy , respectively .
if the extinction could be causing the d1 source to be hidden , then one would expect that both sources would be equally hidden in both bands , unless the p@xmath2 flux at the location of d1 is actually coming from something other than d1 . if d1 were intrinsically as bright as d2 at 1.90@xmath1 m , but being hidden by extinction , then the ratio of their fluxes gives an equivalent magnitude difference of 1.9 mag due to extinction .
we can make the same analysis for the hypothetical d1 source using irac measurements .
if a point source is present at d1 , then its brightness must be at least 40 times lower than the source at d2 .
otherwise , we would detected the irac source looking slightly extended , and the position would be closer to the midpoint between d1 and d2 .
that means that a source at d1 must be at least 1.5 - 2.5 mag fainter than the source at d2 . in order to characterize the properties and the evolutionary phase of the stellar source responsible for the excitation of ionized material in source d
, we obtained the spectral energy distributions ( seds ) of the compact source closest to d2 using measurements from 2mass , irac and pa@xmath2 and 1.90@xmath1 m of nicmos / hst from 1.24@xmath1 m to 8@xmath1 m .
the background subtracted gaussian fitted position of the bright 1.90@xmath1 m source gives the following coordinates @xmath59 .
similarly , the peak radio position d2 at 43 ghz is at @xmath60 .
the 1.9@xmath1 m fitted position gives an @xmath2 , @xmath61 offset of @xmath62 and 0.36@xmath45 , respectively , southeast of the peak radio emission in d2 .
given the random pointing errors and absolute uncertainty in the 1.9@xmath1 m positions @xmath63 , it is likely that the 1.90@xmath1 m source could lie between d1 and d2 but closer to d2 by @xmath64 than to d1 . due to the poor resolution of the 12.8@xmath1 m continuum data
, we can not identify the relative position of the continuum emission with respect to sources d1 and d2 .
lastly , the irac galcen and glimpse ii catalogs identify the position of the irac source as @xmath65 . with a 0.2@xmath45 and 0.3 " uncertainty , respectively .
the centroid of the irac source is shown as a triangle with an uncertainty of 0.2@xmath45 in figure 3a .
this makes it very likely that the irac source corresponds to d2 , and not d1 .
d1 and d2 could have not been separately resolved by irac , but if there were two sources of equal brightness , the point source would have appeared slightly more blurred and more centrally located than observed . given the positional uncertainty ,
we ca nt rule out the possibility of a single source that is centrally located between d1 and d2 , but an association with d2 is most likely . here
, we name source d or g359.96 - 0.08 interchangeably between the central star and the peak of the radio emission .
figure 4 shows irac images of the sgr a east a - d regions at 3.6 , 4.5 , 5.8 , and 8 @xmath66 with sgr a d source being saturated at 8 @xmath66 @xcite .
the associated infrared dark cloud is best seen to the west of the hii regions where submm emission is most prominent , as seen in figure 1 .
d1 and d2 could have not been separately resolved by irac , but if there were two sources of equal brightness , the point source would appear slightly more blurred than observed .
irac data indicate that the peak of the point source does seem to match closer to d2 than d1 .
we used lower limits to the saturated irac fluxes in the sed fitting .
the sed of compact component of d2 is analyzed by comparing to a set of seds produced by a large grid of yso models @xcite .
a linear regression fitter was used to find all seds from the grid of models that are fit with the assumption that the difference between the @xmath67 per datapoint and the @xmath67 per datapoint of the best fit is less than three .
the physical parameters corresponding to those models are averaged along with their standard deviations .
figure 5 shows the fit to the sed of the star associated with source d or g359.96 - 0.08 , and table 3 gives the mean values of visual extinction , stellar mass , luminosity and temperature based on data from 2mass , irac , 12.8@xmath1 m continuum , and upper limits at 450 and 850@xmath1 m .
these physical characteristics are consistent with the assertion that the central star of g359.96 - 0.08 is a hot , massive star .
this star ionizes the surrounding gas , producing thermal radio continuum with a total integrated flux of 81 and 89 mjy at 43.3 and 4.87 ghz , respectively . from the radio continuum and
sed fitting using infrared continuum emission , it appears that the central star is of spectral type ranging between @xmath0b0 and o9 with a luminosity v class @xcite can account for the ionization of the gas in source d , including both d1 and d2 .
lrrrrrrrrr d2 & 42.9@xmath68 & [email protected] & 8.1@xmath69e+04 & 3.6@xmath70e+04 + figure 6 shows four channel maps of [ neii ] line emission , each integrated over a velocity range of 15 . as is true of the radio continuum distribution ,
the [ ne ii ] emission is symmetric about a north - south line between d1 and d2 , but only if the velocities are reflected about v@xmath15 @xmath0 47 . in an alternative display , the [ ne ii ] data cube
is displayed in figure 7 as a series of position - velocity diagrams .
the panels in figure 7 are cuts in right ascension at six different declinations , separated by 1.1 . in the cut passing at declination of @xmath71 , south of the peak emission ,
the line emission , as shown in the third panel from left , is narrow , both spectrally and spatially .
the fourth panel from left shows clearly two blended peaks , one at 17@[email protected]@xmath74 , and one at 51.56@xmath74 . following the naming used for the radio peaks
, we will refer to them at d1 and d2 .
d1 peaks at v@xmath15 = 48 , and has a red shoulder extending to 60 .
d2 peaks at 45 , and has a blue shoulder extending to 35 .
going to the north , the spatial and spectral separation between d1 and d2 continues to increase , and the red and blue shoulders become more prominent . in the fifth panel from left , about 1.5 north of the peak , at @xmath75 , the emission pattern forms an almost complete ring in the position - velocity diagram
all of the position - velocity diagrams show the same symmetry seen in the channel maps , being unchanged under a reflection through ra= 17@[email protected]@xmath74 , and 47 .
the strong symmetry of the ionized gas kinematics about a line between d1 and d2 makes it even less likely that these peaks could be ionized by two stars or that they could be ionized by a star centered on d2 .
it is possible that the ionizing star lies on the line of symmetry .
however , the position of the star at 1.90@xmath1 m shows that the source lies closer to d2 than d1 . unlike the compact hii source d
, sources a , b , and c appear rather similar , both morphologically , as seen in figures 2a , b , and kinematically , in the [ ne ii ] data cube .
morphologically , source a appears to have cometary structure similar to that of g29.96 - 0.02 ( e.g. * ? ? ?
* ) . to test this suggestion , position - velocity diagrams of ra cuts through source a ( as well as b and probably c )
figure 8a shows the position - velocity diagrams for sources a , b and c. sources b and c appear to have smaller extents than source a both spatially and kinematically .
sources a and b appear to have cometary structure whereas source c is not resolved .
the position - velocity diagrams of ra cuts through hii source a at five different declinations are shown in figure 8b .
the cuts along the northern and southern rims of the shell are narrower spectrally than the cuts passing through the center of the hii region , as shown in middle panels .
the broad lines in the eastern rim of the shell and double lines near the center of the shell show a pattern that is very similar to that seen in many uc hii regions ( e.g. * ? ? ?
the characteristic position - velocity diagrams of several types of compact hii regions shown in figs .
45 - 48 of @xcite , show that the kinematics of source a is clearly cometary .
position - velocity diagrams of all three sources , with 3 wide cuts , are shown in figure 9 .
the position velocity ( pv ) diagram for source b is similar to that for a , but shows a smaller extent in both position and velocity , so is not so well resolved .
source c may be similar , but is even less well resolved .
the pv diagrams of figure 8 match bow shocks in which the stellar wind of an o star sweeps the ionized gas into a thin , paraboloidal shell .
the position - velocity diagrams of sources b and c are very similar to that of source a. most compact hii regions with recognizable kinematics observed in [ ne ii ] fall into the latter category , with the hii regions lying on the near sides of their molecular clouds , presumably because extinction through molecular clouds prevents observation of [ ne ii ] emission from those lying on the far sides .
hii source a ( as well as b and probably c ) is unusual in that the orientation of its position - velocity diagram indicates that its head is tipped toward us .
if it lies on the surface of a molecular cloud with its tail pointing out of the cloud , it must lie on the back side .
this is somewhat surprising given the small extinction to the source .
in addition , its velocity indicates that the star is moving relative to the 50 molecular cloud .
in fact , its kinematics matches very well that expected for ionized gas in a bow shock around an o star moving through molecular gas ( see * ? ? ? * ; * ? ? ? * ; * ? ? ?
the star is moving to the east an toward us at an angle of roughly 30 .
perhaps this should nt be surprising , given that this is the most commonly cited model for a cometary hii region , but it is not what was observed by zhu et al . , who found that most cometary hii regions are actually pressure - driven flows around stationary stars near the surfaces of molecular clouds .
it seems likely that the turbulent environment near the galactic center with high density gradient in molecular gas resulted in stars which rather quickly moved out of their natal clouds and are now drifting through lower density ( and so lower extinction ) molecular gas .
the ionizing stars for a , b , and c are not evident in irac data ( see fig .
if present , they are not noticeably brighter than typical fore / background stars , nor are they located symmetrically at the centers of a - c . in stark contrast , d has an extremely bright point source .
this difference alone indicates a significant difference between sources d and a - c .
additional difference between d and a - c hii regions comes from msx observations .
source d has the extremely bright point source that is increasingly more prominent with increasing wavelength .
however , the relative brightness of d drops noticeably compared to a - c when msx data at 8 and 24@xmath1 m are used .
this suggests a lower dust mass for d , or that the bulk of the dust is warmer .
the radio measurements indicate that d1 is optically thin at 5 - 44 ghz , although the slight increase in flux from d2 at higher frequencies suggests that it may be partially optically thick at the very core .
the 1.9@xmath1 m measurements also indicate clearly a bright stellar source near the peak of the radio emission of d2 . as the emission is predominantly optically thin in radio wavelengths , so adopting a distance 8kpc and an ionized gas temperature of 8000k , the total flux of 81mjy at 43.3ghz yields an estimate of the volume emission measure @xmath76@xmath77 .
the associated hydrogen recombination rate requires an ionization rate @xmath78s@xmath7 , where @xmath79 is the hydrogen recombination coefficient into all levels but the ground state . as some ionizing photons might escape the immediate surroundings ,
be absorbed by dust or be absorbed very close to the exciting star , where free - free optical depth may be large , this estimate is a lower limit on the ionizing photon production rate by the central source .
the parameters of the sed of this source suggest that the stellar source g359.23 - 0.063 is associated with a 25 star ( see table 3 ) and 8@xmath5 .
these parameters compare well with the @xmath80s@xmath7 emitted by a o9/b0.5 star @xcite .
the full extent of the source in the ns direction is about 18@xmath45 .
assuming a source size of 9and an expansion velocity of @xmath81 give an age of @xmath82 years ( this age estimate could be a lower limit if the motion is due to flows along the surfaces of the cloud . ) this presents the earliest phase of massive star formation in the 50 cloud .
the estimate for the mass of stars exciting sources a - c can not be made as the stellar source is not identified in these hii regions . however ,
if we assume that the a - c sources have similar masses to that of d , we can make an estimate of the total mass of stars in this regions . to compute the total cluster mass of sgr a east ,
we adopt a standard broken power - law form of the imf @xcite and assume 100 in stars above 20 .
this corresponds to a total stellar mass above 0.5 of 940 associated with a group of young stars in sgr a east .
assuming that the typical age of the cluster is @xmath83 years , we estimate star formation rate of 0.01 to 0.001 yr@xmath7 .
we can also make an estimate of the total mass of molecular gas that went into forming stars .
we used 850 @xmath1 m images of figure 1 to estimate the depleted gas mass by converting submillimeter flux to the mass of molecular gas .
we estimate a subtracted flux of 15 jy at 850@xmath1 m depleted from the region where the group of hii regions is distributed .
we assumed that dust emission from the 50 cloud was uniformly distributed inside and outside the hii regions prior the formation of stars . for a dust temperature of 20 k and a conversion factor from 850@xmath1 m to gas mass @xcite
, we find a rough total gas mass of 7.7@xmath84 that dissipated and converted into 940 of stars .
the near - infrared observations show that despite the ionized gas morphology suggesting a pair of compact hii region , it is probably a single source that is exciting both d1 and d2 .
presumably d1 is produced by irradiation of a dense clump of gas distributed within a cavity that has been partly evacuated of lower density gas by the stellar radiation field . in this scenario ,
the clump has an angular diameter of @xmath85au and lies at a projected distance of @xmath86au from d2 , and so intercepts approximately 10% of the ionizing photons escaping the d1 region .
the observed 30mjy flux at 43ghz requires a total ionization rate at the clump surface of @xmath87s@xmath7 , implying that the exciting source is emitting ionizing photons at a rate @xmath88s@xmath7 .
we now attempt to construct a model to explain the observations of source d. in particular , we need to explain the spatial distribution of the ionized gas and infrared continuum emission , as well as the motions of the ionized gas .
the ionized gas is distributed between two sources , d1 and d2 , each of which is elongated in the ns direction .
it is nearly symmetric about a ns line passing between d1 and d2 , but with the eastern peak , d1 , lying somewhat north of the western peak , d2 , with a p.a .
the infrared continuum , at both 1.9@xmath1 m and 12.8@xmath1 m shows a single peak that lies between the ionized gas peaks , but probably closer to d2 .
the kinematics of the ionized gas can be described by two probably related structures .
a pair of spectrally broad features extend to the north from the d1 and d2 peaks .
the eastern feature extends in velocity from near the molecular cloud velocity redward by about 30 .
the western feature extends about 30 blueward of the molecular cloud .
lower velocity emission is seen both north and south of the peaks , broadening both spectrally and spatially going to the north .
the two broad - lined emission features seem most naturally explained by a bipolar jet pair , probably interacting with the wall of a cavity
. the jet could originate from a young star surrounded by a disk , lying between d1 and d2 .
the disk axis is tipped at a p.a .
@xmath16 70 , with the west side tipped toward us , so that the western jet is blueshifted .
the fact that the jets are not seen extending to the east and west of d1 and d2 indicates that the observed [ ne ii ] emission is not from the jets themselves , but from the region of interaction between the jets and a surrounding wall .
in fact , the bright emission peaks could be simply regions of a wall which are illuminated by ionizing radiation , but the width of the emission lines indicates that jets are needed .
the offset between the position of the 1.9 @xmath1 m continuum and the position of the axis of the symmetry of d1/d2 can also be explained by inhomogeneity of gaseous material in the vicinity of the star and/or the motion of the star .
figure 9 shows a schematic diagram of the relative location of d1 , d2 and the central massive star with a flaring disk .
the fact that d1 is redshifted and d2 is blueshifted could be explained if the disk is seen at an angle between edge - on and face - on , with the west side of the disk tipped toward us .
the lack of symmetry in this diagram is likely the result of density gradient surrounding the star .
the 50 molecular cloud lies mainly to the west of the central star .
high mass stars efficiently photoevaporate their circumstellar disks on timescales of 10@xmath89 10@xmath90 years .
the piling up of the resulting disk wind as it expands against the surrounding medium could be responsible for the @xmath91 of ionized material we identify as sources d1 and d2 .
the semianalytic photoevaporated disk - wind model of @xcite applied to a 28 m star yields a mass flux @xmath92myr@xmath7 , indicating an age of @xmath93 years .
the momentum of the material would be maintained by contributions from the momentum fluxes @xmath94mkms@xmath7 from each of the disk wind ( @xmath95,kms@xmath7 ) @xcite and stellar wind ( @xmath96myr@xmath7 and v @xmath0 1000 ) .
another potential contribution could arise from a fraction of the total momentum flux @xmath97l@xmath98myr@xmath7kms@xmath7 in the photons emitted by the central star .
the extensions to the north could trace the regions of interaction of the jets with the cavity wall if the star is moving to the south , although it is not apparent why broad - lined emission would persist after the jets pass by .
an explanation is also needed for the probable offset of the infrared continuum emission toward d2 .
one possibility is that the disk that collimates the jets is close enough to edge - on that it prevents a direct view of the central star even in the infrared .
the near infrared continuum radiation could be scattered into our line of sight by the material that forms the ionized gas peak d2 . since d1 lies on the far side of the disk , backscattering ( which would be less efficient )
would be required for it d1 to be seen in scattered light .
it may also be obscured by the disk if the disk extends out far enough to cover it .
the 12.8@xmath1 m continuum is unlikely to be scattered , but it could also be affected by extinction , and it is not so convincingly offset from the center of d as is the 1.9@xmath1 m emission .
the spectrally narrow emission extending both north and south of the d peaks is even more difficult to explain than the broad - lined emission .
the ring - like pattern seen in the position - velocity diagrams in figure 7 could be due to gas in each cut lying on an expanding ring .
the fact that the pv rings increase in spatial and velocity extent going to the north would indicate that the three - dimensional ionized gas distribution is on the surface of an expanding cone , with the expansion speed increasing to the north .
alternatively , the gas could be flowing along the surface of a cone if it is accelerating so as to make its speed increase linearly with declination .
however , we can not propose a physical model that would cause either of these flow patterns .
we know of no other case of an expanding cone like what we suggest or any reason a conical wall would expand in this way .
a flow along the surface of a cone seems more natural , but all cometary hii regions we have observed , with sources a - c being representative , are paraboloidal rather than conical , with much larger opening angles that in d. they also have most of their acceleration in a small distance from their vertexes .
( a constant @xmath99 would result in @xmath100 , or @xmath101 , rather than the observed constant @xmath102 . )
it is especially puzzling that the ionized cone has its vertex 3 south of d , so presumably leading the ionizing star , which we think is moving in that direction .
we have to conclude that we are unable to propose a consistent model for the north - south velocity of ionized gas associated with source d. the narrow - lined gas appears to be distributed on the surface of a wedge or cone with an opening angle @xmath020 , with its vertex @xmath03south of the star
. why are cometary hii sources a - c oriented head - on , with velocity offsets indicating that their ionizing stars are moving through the 50 molecular cloud , whereas most cometary hii regions seen in [ ne ii ] are seen tail - on with little stellar motion ?
the answer may lie in the circumstances of their formation .
the location of the sgr a east hii regions near the edge of the sgr a east supernova remnant suggests that the ionizing stars may have formed as a result of compression of molecular gas by the snr .
if the stars formed in swept - up gas , they would have formed with the velocity of the gas .
after their formation , the compressed shell would have slowed as more gas was swept up , and the stars could have drifted out of the compressed gas .
the stars may now be moving through gas of low enough density that extinction does not affect the [ ne ii ] emission , as it would if they were still inside of a dense molecular cloud .
from the orientation of its broken shell , it appears that the ionizing star of hii source a is moving to the east , as well as toward us . from the offset of its position - velocity diagram ,
its motion relative to the 50 molecular cloud is @xmath030 .
the ionizing stars of sources b and c have similar motions . in particular , the position - velocity diagrams indicate that all three sources are moving toward us , while source b is moving to the east and source c is moving to the northeast .
the @xmath103kms@xmath7 velocity width of the [ neii ] line is consistent with other ultra - compact hii regions ( e.g. * ? ? ?
the characterization of a cluster of hii regions sgr a a - d may not be correct , as these hii regions appear to be unbound due to their large physical separation , signifying isolated star formation near the galactic center . again
, the site of on - going star formation is consistent with sites of massive star formation elsewhere ( e.g. * ? ? ?
we have presented multi - wavelength observations of a chain of hii regions located at the eastern edge of the sgr a east supernova remnant and the 50 molecular cloud .
these hii regions show the youngest star formation activity closest to the galactic center .
the youngest member of sgr a east hii regions g359.956 - 0.08 or the sgr a d source is estimated to have an age of roughly 1.5@xmath5 years and consists of two elongated ionized features .
we identified the central star responsible for ionizing both components of the d source .
the sed fit of the central star indicates a mass of 25 and a luminosity of 8@xmath5 .
the kinematics of ionized gas show a an e - w red and blue shifted flow with respect to the central star and a complex asymmetric velocity structure in the n - s direction .
we presented a simple model in which the central star is surrounded by a disk constraining the flow of ionized wind . in this picture
, the uv radiation from the central hot star may generate an outflow from the disk , sweeping up the ism gas and forming the two components of ionized gas d1 and d2 .
given that high density molecular material is distributed to the west , the ionized feature d2 is expected to lie closer to the central star .
future high spectral resolution observations of ionized gas and detailed modeling of the complex n - s velocity structure should explain better the kinematics of ionized gas presented here .
this work is partially supported by grants ast-0807400 ( to fyz ) and ast-0607312 ( to jhl ) from the national science foundation .
this research has made use of the nasa/ ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
99 arendt , r. g. ; stolovy , s. r. ; ramrez , s. v. ; sellgren , k. ; cotera , a. s. ; law , c. j. ; yusef - zadeh , f. ; smith , h. a. ; gezari , d. y. 2008 , , 682 , 384 arthur , s. j. & hoare , m. g. 2006 , , 165 , 283 bartko , h. , martins , f. , fritz , t. k. , genzel , r. , levin , y. , perets , h. b. , paumard , t. , nayakshin , s. , gerhard , o. , alexander , t. et al . 2009 , , 697 , 1741 benjamin , r. a. , et al .
2003 , , 115 , 953 cotera , a. s. , simpson , j. p. , erickson , e. f. , colgan , s. w. j. , burton , m. g. , allen , d. a. 1999 , , 510 , 747 depree , c. g. , goss , w. m. & gaume , r. a. 1998 , , 500 , 847 drew , j. e. , proga , d. & stone , j. m. 1998 , , 296 , l6 fey , alan l. ; gaume , ralph a. ; claussen , mark j. ; vrba , frederick j. 1995 , , 453 , 308f figer , d. f. , najarro , f. , gilmore , d. , morris , m. , kim , s. s. , serabyn , e. , mclean , i. s. , gilbert , a. m. , graham , j. r. , larkin , j. e. , levenson , n. a. , & teplitz , h. i. 2002 , 581 , 258 garay , g. & lizano , s. 1999 , , 111 , 1049 , a. m. , salim , s. , weinberg , n. n. , lu , j. r. , do , t. , dunn , j. k. , matthews , k. , morris , m. r. , yelda , s. , becklin , e. e. , kremenek , t. , milosavljevic , m. , & naiman , j. 2008 , , 689 , 1044 , s. , eisenhauer , f. , quataert , e. , genzel , r. , paumard , t. , trippe , s. , ott , t. , abuter , r. , eckart , a. , lagage , p. o. , lehnert , m. d. , tacconi , l. j. , & martins , f. 2006 , , 640 , l163 goss , w. m. , schwarz , u. j. , van gorkom , j. h. & ekers , r. d. 1985 , , 215 , 69p hollenbach , d. , johnstone , d. , lizano , s. & shu , f. 1994 , , 428 , 654 hoare , m. g. 2006 , , 649 , 856 hoare , m. g. , kurtz , s. e. , lizano , s. , keto , e. & hofner , p. 2007 , protostars and planets v , p181 , p. 2001
, , 322 , 231 lacy , j. h. , richter , m. j. , greathouse , t. k. , jaffe , d. t. & zhu , q. 2002 , , 114 , 153 morris , m. & serabyn , e. 1996 , , 34 , 645 mckee , c. f. & ostriker , e. c. 2007 , , 45 , 565 panagia , n. 1973 , , 78 , 929 , d. , richer , j. s. , greaves , j. s. , holland , w. s. , jenness , t. , lasenby , a. n. , white , g. j. , matthews , h. e. , ward - thompson , d. , dent , w. r. f. , zylka , r. , mezger , p. , hasegawa , t. , oka , t. , omont , a. , & gilmore , g. 2000 , , 545 , l121 ramirez , s. v. , arendt , r. g. , sellgren , k. , stolovy , s. r. , cotera , a. , smith , h. a. , & yusef - zadeh , f. 2008 , , 175 , 147 , m. j. , & brunthaler , a. 2004 , , 616 , 872 , t. p. , whitney , b. a. , indebetouw , r. , & wood , k. 2007 , , 169 , 328 shu , f. h. & adams , f. c. & lizano , s. 1987 , , 25 , 23 serabyn , e. , lacy , j. h. , & achtermann , j. m. 1992 , , 395 , 166 stolovy , s. , et al .
2006 , journal of physics conference series , 54 , 176 tsuboi , m. , miyazaki , a. & okumura , s. k. 2009 , , 61 , 29 vacca , w. d. , garmany , c. d. & shull , j. m. 1996 , , 460 , 914 wang , q. d. , dong , h. , cotera , a. , stolovy , s. , morris , m. , lang , c. c. , muno , m. p. , schneider , g. & calzetti , d. 2010 , , 402 , 895 wardle , m. & yusef - zadeh , f. 2008 , , 683 , l37 , b. a. , wood , k. , bjorkman , j. e. , & cohen , m. 2003 , , 598 , 1079 wilkin , f. p. 1996
, , 459 , l31 yusef - zadeh , f. & morris , m. 1987 , , 320 , 545 yusef - zadeh , f. , telesco , c. m. & decher , r. 1989 , iau symp .
136 , ed : mark morris , kluwer academic publishers , dordrecht , p.287 yusef - zadeh , f. & wardle , m. 1993 , , 405 , 584 zhu , q - f . ; lacy , j. h. , jaffe , d. t. , richter , m. j. & greathouse , t. k. 2008 , , 177 , 584 zinnecker , h. & yorke , h. w. 2007 , , 45 , 481 m emission from the sgr a east hii region are superimposed on a 20 cm continuum image ( black ) and are represented at levels 32 , 34 , ... , 50 jy beam@xmath7 ( beam size is 8@xmath104 ) . *
( b ) right * similar to ( a ) except that contours of 850@xmath1 m emission are presented at levels ( 32 , 34 , .. , 46 ) @xmath105 jy beam@xmath7 ( beam size is 15@xmath106 ) . the greyscale 20 cm continuum image ( in black )
has a resolution of 3.5 and 2.9 ( pa=26@xmath107 ) .
the 450@xmath1 m and 850@xmath1 m emission trace dust emission from the 50 molecular cloud gmc m-0.02 - 0.07 .
, title="fig : " ] m emission from the sgr a east hii region are superimposed on a 20 cm continuum image ( black ) and are represented at levels 32 , 34 , ... , 50 jy beam@xmath7 ( beam size is 8@xmath104 ) . *
( b ) right * similar to ( a ) except that contours of 850@xmath1 m emission are presented at levels ( 32 , 34 , .. , 46 ) @xmath105 jy beam@xmath7 ( beam size is 15@xmath106 ) . the greyscale 20 cm continuum image ( in black )
has a resolution of 3.5 and 2.9 ( pa=26@xmath107 ) .
the 450@xmath1 m and 850@xmath1 m emission trace dust emission from the 50 molecular cloud gmc m-0.02 - 0.07 .
, title="fig : " ] ) .
the top three hii sources a to c are resolved having shell - like structures whereas the most compact hii source d shows two compact sources elongated in the north - south direction . *
( b ) top right * the same as ( a ) except that the map is based on integrated [ neii ] line emission over a velocity range of v@xmath15 = 10 to 70 corresponding to spectral interval of 0.158 cm@xmath7 .
contours of the line emission are set quadratically at ( 9 , 16 , 25 , ... ) @xmath109 erg s@xmath7 @xmath110 sr@xmath7 . *
( c ) bottom left * contour map of source d before and after deconvolution averaged over a 70 km s@xmath7 interval centered at 50 km s@xmath7 with contour intervals 0.02 erg s@xmath7 @xmath110 sr@xmath7 . *
( d ) bottom right * the line and continuum maps are shown in the left and right panels with contour intervals of 2.6@xmath111 and 3@xmath111 erg s@xmath7 @xmath110 sr@xmath7 , respectively . the line map is the average over a 200 km / s interval centered on the line . ,
title="fig : " ] ) .
the top three hii sources a to c are resolved having shell - like structures whereas the most compact hii source d shows two compact sources elongated in the north - south direction . *
( b ) top right * the same as ( a ) except that the map is based on integrated [ neii ] line emission over a velocity range of v@xmath15 = 10 to 70 corresponding to spectral interval of 0.158 cm@xmath7 .
contours of the line emission are set quadratically at ( 9 , 16 , 25 , ... ) @xmath109 erg s@xmath7 @xmath110 sr@xmath7 . * ( c ) bottom left * contour map of source d before and after deconvolution averaged over a 70 km s@xmath7 interval centered at 50 km s@xmath7 with contour intervals 0.02 erg s@xmath7 @xmath110 sr@xmath7 . *
( d ) bottom right * the line and continuum maps are shown in the left and right panels with contour intervals of 2.6@xmath111 and 3@xmath111 erg s@xmath7 @xmath110 sr@xmath7 , respectively . the line map is the average over a 200 km / s interval centered on the line . ,
title="fig : " ] + ) . the top three hii sources a to c
are resolved having shell - like structures whereas the most compact hii source d shows two compact sources elongated in the north - south direction . *
( b ) top right * the same as ( a ) except that the map is based on integrated [ neii ] line emission over a velocity range of v@xmath15 = 10 to 70 corresponding to spectral interval of 0.158 cm@xmath7 .
contours of the line emission are set quadratically at ( 9 , 16 , 25 , ... ) @xmath109 erg s@xmath7 @xmath110 sr@xmath7 . *
( c ) bottom left * contour map of source d before and after deconvolution averaged over a 70 km s@xmath7 interval centered at 50 km s@xmath7 with contour intervals 0.02 erg s@xmath7 @xmath110 sr@xmath7 . *
( d ) bottom right * the line and continuum maps are shown in the left and right panels with contour intervals of 2.6@xmath111 and 3@xmath111 erg s@xmath7 @xmath110 sr@xmath7 , respectively . the line map is the average over a 200 km / s interval centered on the line . ,
title="fig : " ] mjy per beam .
the cross and the triangle show the peak positions of the 1.90@xmath1 m emission and the catalogued position of irac source with an uncertainty of 0.1@xmath45 to @xmath63 , respectively .
both sources lie to the southeast of the peak radio emission from d2 . *
( b ) right * a grayscale image of source d at 43.3 ghz with a spatial resolution of 0.58 and 0.47(pa=8@xmath107 ) .
contours are set at -1 , 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 mjy per beam .
, title="fig : " ] mjy per beam .
the cross and the triangle show the peak positions of the 1.90@xmath1 m emission and the catalogued position of irac source with an uncertainty of 0.1@xmath45 to @xmath63 , respectively .
both sources lie to the southeast of the peak radio emission from d2 . *
( b ) right * a grayscale image of source d at 43.3 ghz with a spatial resolution of 0.58 and 0.47(pa=8@xmath107 ) .
contours are set at -1 , 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 mjy per beam .
, title="fig : " ] ( left to right , top to bottom ) .
the pixel size is @xmath112 .
bright sources at 5.8 and 8 @xmath66 exhibit detector artifacts @xmath113 and @xmath114 to their south .
this is especially distinct for sgr a east d , which is the brightest source in the field and is obviously saturated at 8 @xmath66 . ]
line intensity are set quadratically at ( 9 , 16 , 25 , ... ) @xmath116 erg s@xmath7 @xmath110 cm sr@xmath7 .
the absolute coordinates of [ neii ] line images are done by comparing radio and [ neii ] images with each other . ] position velocity diagrams of sgr a east a source showing the velocity cuts at constant declination of ( -29@xmath117 ) of 31.2 29.1 26.9 24.8 22.6 .
contour levels are set quadratically at ( 9 , 16 , 25 , ... ) @xmath118 erg s@xmath7 @xmath110 cm sr@xmath7 . ] | a group of four compact hii regions associated with the well - known 50 molecular cloud is the closest site of on - going star formation to the dynamical center of the galaxy , at a projected distance of @xmath06 pc .
we present a study of ionized gas based on the [ neii ] ( 12.8@xmath1 m ) line , as well as multi - frequency radio continuum , hst pa@xmath2 and spitzer irac observations of the most compact member of the hii group , sgr a east hii d. the radio continuum image at 6 cm shows that this source breaks up into two equally bright ionized features , d1 and d2 .
the sed of the d source is consistent with it being due to a 25@xmath3 star with a luminosity of @xmath4 .
the inferred mass , effective temperature of the uv source and the ionization rate are compatible with a young o9-b0 star .
the ionized features d1 and d2 are considered to be ionized by uv radiation collimated by an accretion disk .
we consider that the central massive star photoevaporates its circumstellar disk on a timescale of 3@xmath5 years giving a mass flux @xmath6myr@xmath7 and producing the ionized material in d1 and d2 expanding in an inhomogeneous medium .
the ionized gas kinematics , as traced by the [ ne ii ] emission , is difficult to interpret , but it could be explained by the interaction of a bipolar jet with surrounding gas along with what appears to to be a conical wall of lower velocity gas .
the other hii regions , sgr a east a - c , have morphologies and kinematics that more closely resemble cometary flows seen in other compact hii regions , where gas moves along a paraboloidal surface formed by the interaction of a stellar wind with a molecular cloud .
km s@xmath7 # 1 10^#1 # 1 # 2 ^#2 u#1
i v |
an epidemic working through a population , cascading electrical power failures , product adoption , and the spread of a mutant gene are all examples of diffusion processes that can happen in multi - agent systems structured as complex networks .
these network processes have been studied in a variety of disciplines , including computer science @xcite , biology @xcite , sociology @xcite , economics @xcite , and physics @xcite .
much existing work in this area is based on pre - existing models in sociology and economics in particular the work of @xcite .
however , recent examinations of social networks both analysis of large data sets and experimental have indicated that there may be additional factors to consider that are not taken into account by these models .
these include the attributes of nodes and edges , competing diffusion processes , and time . in this paper ,
we outline seven design criteria ( section [ sec : criteria ] ) for such a framework and introduce @xmath0 ( section [ prelimsec ] ) , which is to the best of our knowledge the first logical language for modeling diffusion in complex networks that meets these criteria .
@xmath0 is a rule - based framework ( inspired by logic programming ) that can richly express how agents adopt or fail to adopt certain behaviors , and how these behaviors cascade through a network .
we also introduce fixed - point based algorithms that allow for the calculation of the result of the diffusion process in section [ fpsec ] .
note that these algorithms are proven not only to be correct , but also to run in polynomial time .
hence , our approach can not only better express many aspects of cascades in complex networks , but it can do so in a reasonable amount of time .
we conclude by discussing applications of @xmath0 in section [ sec : learn ] .
_ proofs of all results stated in this paper can be found in the appendix .
_ we begin by identifying a set of criteria that we believe a framework for reasoning about cascades in complex networks should satisfy . * 1 .
multiply labeled and weighted nodes and edges . *
many existing frameworks for studying diffusion in complex networks assume that there is only one type of vertex that may become `` active '' @xcite or may `` mutate '' @xcite and only one possible relationship between nodes . in reality ,
nodes and edges often have different properties .
for instance , labels on edges can be used to differentiate between strong and weak ties ( edge types ) a concept that is well studied @xcite .
recently , such attributes of nodes have been shown to impact influence in a network @xcite . *
2 . explicit representation of time . *
most work in the literature assumes static models , with the exception of the recent developments in @xcite , which assume the existence of a timestamped log referring to actions taken in the network in order to learn how nodes influence each other . though @xcite tackles the problem of predicting the time at which a certain node will take an action , the authors make several simplifying assumptions such as monotonicity of probability functions , probabilistic independence , sub - modularity and , most importantly for this criterion , a modeling of time solely based on temporal decay of influence .
we seek a richer model of temporal relationships between conditions in the network structure , the current state of the cascades in process , and how influence propagates . *
non - markovian temporal relationships . *
apart from time being explicitly represented , the temporal dependencies should be able to span multiple units of time .
hence , the `` memoryless '' mode of a standard markov process , where only the information of the current state is required , is insufficient . here , we strive to create a framework where dependencies can be from other earlier time steps .
this issue has been previously studied with respect to more general logic programming frameworks such as @xcite , but to our knowledge has not been applied to social networks . *
representation of uncertainty . * as in practice it is not always possible to judge the attributes of all individuals in a network , an element of uncertainty must be included .
however , in connection with point 7 , this should not be at the expense of tractability .
for instance , the probabilistic models of @xcite are normally addressed with simulation ( and hence do not scale well ) as the computation of the expected number of activated nodes is a @xmath1-hard problem @xcite . *
competing cascades . * often , in real - world situations there will be competing cascading processes .
for example , in evolutionary graph theory @xcite , `` mutants '' and `` residents '' compete for nodes in the network the success of one hinges on the failure of the other .
. non - monotonic cascades . *
in much existing work on cascades in complex networks , the number of nodes attaining a certain property at each time step can only increase .
however , if we allow for competing cascades in the same model , we can not have such a strong restriction as the success of one cascade may come at the expense of another . *
7 . tractability . * the social networks of interest in today s data mining problems often have millions of nodes .
it is reasonable to expect that soon billion - node networks will be commonplace .
any framework for dealing with these problems must be solvable in a reasonable amount of time and offer areas for practical improvement for further scalability .
the above criteria can be summarized as the desire to design the most expressive language for network cascades possible while still allowing computation of the outcome of a diffusion process to be completed in a tractable amount of time . as a comparison ,
let us briefly describe some relevant related work .
perhaps the best known general model for representing diffusion in complex networks is the independent cascade / linear threshold ( ic / lt ) model of @xcite .
however , although this framework was shown to be capable of expressing a wide variety of sociological models , it assumes the markov property and does not allow for the representation of multiple attributes on vertices and edges .
a more recent framework , social network optimization problems ( snops ) @xcite uses logic programming to allow for the representation of attributes , but this framework does not allow for competing processes or non - monotonic cascades . a related logic programming framework , competitive diffusion ( cd ) @xcite allows for competitive diffusion and non - monotonic processes but does not explicitly represent time and also makes markovian assumptions .
further , we also note that the semantics of cd yields a `` most probable interpretation '' that is not a unique solution .
hence , a given model in that framework can lead to multiple and possibly contradictory , outcomes to a cascade ( this problem is avoided in @xmath0 ) .
another popular class of models is evolutionary graph theory ( egt ) @xcite , which is highly related to the voter model ( vm ) @xcite .
although this framework allows for competing processes and non - monotonic diffusion , it also makes markovian assumptions while not explicitly representing time .
further , determining the outcome of a cascade in those models is np - hard , while determining the outcome in @xmath0 can be accomplished in polynomial time .
table [ rwtab ] lists how these models compare to @xmath0 when considering our design criteria . [
cols="<,^,^,^,^,^,^",options="header " , ] we define a @xmath0 interpretation as follows . a @xmath0 interpretation @xmath2 is a mapping of natural numbers in the interval @xmath3 $ ] to network interpretations , i.e. , @xmath4 . let @xmath5 be the set of all possible interpretations .
first , we define what it means for an interpretation to satisfy a fact and a rule .
an interpretation @xmath2 * satisfies * @xmath0 fact @xmath6 $ ] , written @xmath7 $ ] , iff @xmath8 $ ] , @xmath9 .
[ satfactex ] consider interpretation @xmath10 , where @xmath11 ( from example [ niex ] ) , and @xmath0 facts @xmath12 and @xmath13 from example [ ex3 ] . in this case ,
@xmath14 and @xmath15 . for non - fluent facts ,
we introduce the notion of strict satisfaction , which enforces the bound in the interpretation to be set to exactly what the fact dictates .
interpretation @xmath2 * strictly satisfies * @xmath0 fact @xmath16 $ ] iff @xmath8 $ ] , @xmath17 . next , we define what it means for an interpretation to satisfy an integrity constraint .
an interpretation @xmath2 * satisfies * integrity constraint @xmath18 iff for all @xmath19 and @xmath20 , @xmath21 .
before we define what it means for an interpretation to satisfy a rule , we require two auxiliary definitions that are used to define the bound enforced on a label by a given rule , and the set of time points that are affected by a rule .
[ bounddef ] for a given rule @xmath22 , node @xmath23 , and network interpretation @xmath24 , @xmath25 @xmath26 where @xmath27 @xmath28 and @xmath29 @xmath30 intuitively , the bound returned by the function depends on the influence function and the number of qualifying and eligible nodes that influence it . for interpretation @xmath2 , node @xmath23 , and
rule @xmath22 , the _ target time set _ of @xmath31 is defined as follows : @xmath32 \ ; | \ ; i(t-\delta t)(v ) \models f\big\}\ ] ] we also extend this definition to a program @xmath33 , for a given @xmath20 and @xmath34 , as follows ; @xmath35 @xmath36 \ ; | \ ; ( \<l,\bnd\>,c):[t_1,t_2 ] \in p\big\}\ ] ] @xmath37 we can now define satisfaction of a rule by an interpretation .
an interpretation @xmath2 * satisfies * a rule @xmath38 iff for all @xmath39 and @xmath40 it holds that @xmath41 [ satrule ] let @xmath10 be the interpretation from example [ satfactex ] .
suppose that @xmath42\ > \in i(1)(5)$ ] . in this case ,
let @xmath44 be equivalent to @xmath10 except that we have @xmath45\ > \in i_2(1)(3)$ ] .
in this case , @xmath46 .
we now define satisfaction of programs , and introduce _
canonical interpretations _ , in which time points that are not `` targets '' retain information from the last time step . for interpretation @xmath2 and program @xmath33 : @xmath2 is a * model * for @xmath33 iff it satisfies all rules , integrity constraints , and fluent facts in that program , strictly satisfies all non - fluent facts in the program , and for all @xmath47 @xmath20 and @xmath48 , @xmath49\ > \in i(c)(t)$ ] .
@xmath2 is a * canonical model * for @xmath33 iff it satisfies all rules , integrity constraints , and fluent facts in @xmath33 , strictly satisfies all non - fluent facts in @xmath33 , and for all @xmath47 @xmath50 and @xmath48 , @xmath49\ > \in i(c)(t)$ ] when @xmath51 and @xmath52 where @xmath53 , otherwise .
[ canonex ] following from previous examples , if we consider interpretation @xmath10 and program @xmath54 , we have that @xmath55\>$ ] must be in @xmath56 in order for @xmath10 to be canonical . in this section
we discuss consistency and entailment in @xmath0 programs , and explore the use of minimal models towards computing answers to these problems .
a @xmath0 program @xmath33 is ( canonically ) consistent iff there exists a ( canonical ) model @xmath2 of @xmath33 . a @xmath0 program @xmath33 ( canonically ) entails @xmath0 fact @xmath57 iff for all ( canonical ) models @xmath2 of @xmath33
, it holds that @xmath58 .
now we define an ordering over models and define the concept of minimal model .
we then show that if we can find a minimal model then we can answer consistency , entailment , and tight entailment queries .
to do so , we first define a pre - order over interpretations .
given interpretations @xmath59 we say @xmath60 if and only if for all @xmath61 if there exists @xmath62 then there must exist @xmath63 s.t .
@xmath64 .
next , we define an equivalence relation for interpretations denoted with @xmath65 ; we will use the notation @xmath66 $ ] for the set of all interpretations equivalent to @xmath2 w.r.t .
this allows us to define a partial ordering .
two interpretations @xmath67 are * equivalent * ( written @xmath68 ) iff for all @xmath69 , @xmath70 iff @xmath71 . given classes of interpretations @xmath66,[i']$ ] that are equivalent w.r.t .
@xmath65 , we say that @xmath66 $ ] precedes @xmath72 $ ] , written @xmath66 \sqsubseteq [ i']$ ] , iff @xmath60 .
the partial ordering is clearly reflexive , antisymmetric , and transitive .
note that we will often use @xmath73 as shorthand for @xmath66 \sqsubseteq [ i']$ ] .
we define two special interpretations , @xmath74 and @xmath75 , such that @xmath76 , @xmath77 and there exists network atom @xmath78 .
clearly , no other interpretation can be below @xmath74 as the @xmath79 $ ] bound on all network atoms for each time step and each component is @xmath80 $ ] ; similarly , no other interpretation is above @xmath75 , since for any interpretation @xmath2 for which there exists @xmath81 where @xmath82 , we have @xmath83
. we can prove ( see the full version of the paper for details ) that with @xmath75 and @xmath74 , @xmath84 is a complete lattice .
we can now arrive at the notion of _ minimal model _ for a @xmath0 program .
given program @xmath33 , the minimal model of @xmath33 is a ( canonical ) interpretation @xmath2 s.t .
@xmath70 and for all ( canonical ) interpretation @xmath85 s.t .
@xmath71 , we have that @xmath73 .
suppose we have some algorithm @xmath86 that takes as input a program @xmath33 and returns an interpretation @xmath2 ( where @xmath2 does not necessarily satisfy @xmath33 ) s.t . for all @xmath85 where @xmath71 , @xmath73 .
it is easy to show that if @xmath87 then @xmath33 is consistent . likewise ,
if @xmath88 then @xmath33 is inconsistent , as all models must then have a tighter weight bound for the network atoms than an invalid interpretation ( hence , making such an interpretation invalid as well ) .
clearly , any such algorithm @xmath86 would provide a sound and complete answer to the consistency problem . likewise
, if we consider the entailment problem , it is easy to show that for fact @xmath89 $ ] , @xmath33 ( canonically ) entails @xmath57 iff the minimal model of @xmath33 ( canonically ) satisfies @xmath57 .
this is because for minimal model @xmath90 of @xmath33 , for any time @xmath91 $ ] , if @xmath92 then there is network atom @xmath93 s.t .
we note that for any other interpretation @xmath2 of @xmath33 with @xmath94 we have that @xmath95 .
hence , having a minimal model allows us to solve any entailment query .
we can think of a minimal model of a @xmath0 program as the outcome of a diffusion process in a multi - agent system modeled as a complex network .
hence , a question such as `` how many agents will adopt the product with a weight of at least @xmath96 in two months ? '' can be easily answered once the minimal model is obtained .
in this section we introduce a fixed - point operator that produces the non - canonical minimal model of a @xmath0 program in polynomial time .
this is followed by an algorithm to find a canonical minimal model also in polynomial time .
first , we introduce three preliminary definitions . for a given @xmath0 program @xmath33 , @xmath20 , @xmath34 , and @xmath97 we define function @xmath98
@xmath99\in p\textit { s.t .
} t \in [ t_1,t_2]}\bnd\ ] ] [ ibound ] for a given @xmath0 program @xmath33 , @xmath20 , @xmath34 , and @xmath97 we define function @xmath100 @xmath101 given @xmath0 program @xmath33 , interpretation @xmath2 , @xmath39 , @xmath34 , and @xmath97 , we define @xmath102 @xmath103 we can now introduce the operator . for a given @xmath0 program @xmath33
, we define the operator @xmath104 as follows : for a given @xmath2 , for each @xmath19 , @xmath105 , and @xmath34 , add @xmath106 to @xmath107 where @xmath108 is defined as : @xmath109 where @xmath110 .
it is easy to show that @xmath111 can be computed in polynomial time ( the proof is in the full version ) .
next , we introduce notation for repeated applications of @xmath111 . given natural number @xmath112 , interpretation @xmath2 , and program @xmath33 , we define @xmath113 , the multiple applications of @xmath111 , as follows : @xmath114 we can prove that the iterated @xmath111 operator converges after a polynomial number of applications : [ gammapolyconverge ] given interpretation @xmath2 and program @xmath33 , there exists a natural number @xmath115 s.t.@xmath116 , and @xmath117 where @xmath118 is the maximum in - degree in the network . for a given vertex @xmath119
, we will use the notation @xmath120 to denote the number of incoming neighbors ( of any edge type ) .
first note that for a given @xmath121 and @xmath34 , a given rule @xmath122 can tighten the bound on a network atom formed with @xmath123 no more than @xmath124 times . at each application of @xmath111
, at least one network atom must tighten .
hence , as there are only @xmath125 tightenings possible , this is also the bound on the number of applications of @xmath111 . in the following ,
we will use the notation @xmath126 to denote the iterated application of @xmath111 after a number of steps sufficient for convergence ; theorem [ gammapolyconverge ] means that we can efficiently compute @xmath126 .
we also note that as a single application of @xmath111 can be computed in polynomial time , this implies that we can find a minimal model of a @xmath0 program in polynomial time .
we now prove the correctness of the operator .
we do this first by proving a key lemma that , when combined with a claim showing that for consistent program @xmath33 , @xmath126 is a model of @xmath33 , tells us that @xmath126 is a minimal model for @xmath33 .
following directly from this , we have that @xmath33 is inconsistent iff @xmath127 .
[ boundlemma ] if @xmath70 and @xmath128 then @xmath129 .
[ minmodelgamma ] if program @xmath33 is consistent then @xmath126 is a minimal model for @xmath33 . these results ,
when taken together , prove that tight entailment and consistency problems for @xmath0 can be solved in polynomial time , which is precisely what we set out to accomplish as part of our desiderata described in section [ sec : criteria ] .
next , we develop an algorithm for the _ canonical _ versions of consistency and tight entailment , and show that we can bound the running time of the algorithm with a polynomial .
we also note that subsequent runs of the convergence of @xmath111 will likely complete quicker in practice , as the initial interpretation is the last interpretation calculated ( cf . line [ recalcline ] ) .
we also show that the interpretation produced by the algorithm is a canonical minimal model .
following from that , a program is inconsistent iff the algorithm returns @xmath75 .
program @xmath33 interpretation @xmath2 [ bigfor ] [ condline ] [ vlloop][insideforbegin ] [ insideforend ] [ recalcline ] [ newcurfree ] @xmath2 algorithm performs no more than @xmath130 calculations of the convergence of @xmath111 .
[ canonsound ] if @xmath33 is consistent , then @xmath131 is the minimal canonical model of @xmath33 .
in this section , we will briefly discuss work in progress on how can be applied in real world settings .
it is widely acknowledged that modeling influence in multi - agent systems ( most usefully modeled as complex networks ) is highly desirable for many practical problems as varied as viral marketing , prevention of drug use , vaccination , and power plant failure .
though programs are a rich model to work with , the acquisition of rules is the principal hurdle to overcome ; this is mainly due to this richness of representation , since for each rule we must provide a set of conditions on the agents being influenced , conditions on their neighbors and their ties to their neighbors , and how capable these neighbors are of influencing them .
a domain expert is likely able to provide important insights into these components , but the best way to obtain these rules is undoubtedly to leverage the presence of large amounts of data in domains like twitter ( with about 340 m messages sent per day , available through public apis ) , facebook ( over 950 m users with more complex information ; not publicly available , but data can be requested through apps ) , and blogging and photo hosting sites such as blogger and flickr ( which have millions of users as well ) . concretely , we have begun working towards this goal by extracting several time - series , multi - attribute network data sets on which to apply @xmath0 . for instance , to study the proliferation of research on different topics , we looked at research on `` niacin '' indexed by thomson reuters web of knowledge ( http://wokinfo.com ) .
this topic was chosen due to its interest to a variety of disciplines , such as medicine , biology , and chemistry ; this gives the data more variety compared with more discipline - specific topics .
we extracted an author - paper bipartite network consisting of @xmath132 papers with @xmath133 authors and @xmath134 edges ( cf .
figure [ fig : papers12 ] ) ; from this data we can easily focus on various kinds of networks ( co - author , citation , etc . ) .
we have also collected attribute and time - series data for this network , as well as the subjects of the papers ; the propagation of these subjects is a good starting point to test methods for the acquisition of @xmath0 rules .
we are harvesting larger datasets from various online social networks .
further details can be found in the full version of the paper . * a proposed learning architecture .
* we are currently developing a learning architecture ( depicted in fig .
[ fig : architecture ] ) based on the use of state - of - the - art data analysis , clustering , and influence learning techniques as building blocks for the acquisition of rules from data sets .
the key question is not just the identification of the best techniques to adopt , but how to adapt them and combine them in such a way as to produce meaningful and useful outputs .
consider the diagram in fig .
[ fig : architecture ] : the data first flows from raw data sources to the _ cluster identification _ component , which has the goal of identifying sets of agents behaving as groups ( for instance , teens influencing other teens of the same sex in the consumption of music , or scientists of a certain field influencing the research topics of others in a related field ) @xcite ; the main output here is a set of conditions on nodes and edges that characterize groups of nodes .
once clusters are identified , the _ influence recognition _ component will make use of both the clusters and the data sources to recognize what kind of influence is present in the system @xcite ; the main output of this component is the influence function to be used in the rules .
the _ rule generation _ component then takes the output of the cluster identification and influence recognition components , along with the raw data ( _ e.g. _ , to analyze time stamps ) and produces rules ; the output of this component is involved in a refinement cycle with experts who can provide feedback on the rules being produced ( such as possible combinations of rules , identification of cases of overfitting , etc . ) .
in this paper , we presented the @xmath0 language for modeling cascades in multi - agent systems organized in the form of complex networks .
we started by establishing seven criteria in the form of desiderata for such a formalism , and proved that @xmath0 meets all of them ; to the best of our knowledge , this has not been accomplished by any previous model in the literature .
we also note that @xmath0 is the first language of its kind to consider network structure in the semantics , potentially opening the door for algorithms that leverage features of network topology in more efficiently answering queries .
our current work involves implementing the algorithms described in this paper , as well as the real - world applications described in section [ sec : learn ] ; though our algorithms have polynomial time complexity , it is likely that further optimizations will be needed in practice to ensure scalability for very large data sets . in the near future
, we shall also explore various types of queries that have been studied in the literature , such as finding agents of maximum influence , identifying agents that cause a cascade to spread more quickly , and identifying agents that can be influenced in order to halt a cascade .
p.s . is supported by the army research office ( project 2gdatxr042 ) .
is supported under ( uk ) epsrc grant ep / j008346/1 proqaw .
the opinions in this paper are those of the authors and do not necessarily reflect the opinions of the funders , the u.s .
military academy , or the u.s .
10 s. aral and d. walker . .
, 337(6092):337341 , 2012 .
m. broecheler , p. shakarian , and v. subrahmanian .
a scalable framework for modeling competitive diffusion in social networks . in _ proc . of socialcom_. ieee , 2010 .
w. chen , c. wang , and y. wang .
scalable influence maximization for prevalent viral marketing in large - scale social networks . in _ proc . of kdd 10
_ , pages 10291038 .
acm , 2010 .
a. goyal , f. bonchi , and l. lakshmanan .
discovering leaders from community actions . in _ proc .
of cikm _ , pages 499508 .
acm , 2008 .
a. goyal , f. bonchi , and l. lakshmanan .
learning influence probabilities in social networks . in _ proc .
of wsdm _ , pages 241250 .
acm , 2010 .
a. goyal , f. bonchi , and l. lakshmanan .
a data - based approach to social influence maximization .
, 5(1):7384 , 2011 .
m. granovetter . .
, 78(6):13601380 , 1973 .
m. granovetter .
threshold models of collective behavior . , 83(6):14201443 , 1978 .
a. jain .
data clustering : 50 years beyond k - means .
, 31(8):651666 , 2010 .
d. kempe , j. kleinberg , and e. tardos . maximizing the spread of influence through a social network . in _ proc . of kdd 03
_ , pages 137146 .
acm , 2003 .
e. lieberman , c. hauert , and m. a. nowak .
evolutionary dynamics on graphs . , 433(7023):312316 , 2005 .
t. c. schelling . .
w.w . norton and co. , 1978 .
p. shakarian , a. parker , g. i. simari , and v. s. subrahmanian .
annotated probabilstic temporal logic .
, 12(2 ) , 2011 .
p. shakarian , v. subrahmanian , and m. l. sapino . .
in _ proc . of iclp ( tech .
comm . ) _ , 2010 .
v. sood , t. antal , and s. redner .
voter models on heterogeneous networks .
, 77(4):041121 , 2008 .
t. warren liao .
clustering of time series data a survey .
, 38(11):18571874 , 2005 .
let @xmath135 be a subset of @xmath5 . we can create @xmath136 as follows .
we build interpretation @xmath85 . for each @xmath137 , let @xmath138 be the least of the set @xmath139 \ >
\in i(t)(c),\<l , [ \ell , u ) \ > \in
i(t)(c ) \}$ ] and @xmath140 be the least of the set @xmath141 \
i(t)(c),\<l , ( \ell , u ) \ > \in i(t)(c ) \}$ ] . then , for each @xmath137 let @xmath142 be the greatest element of the set @xmath143 \ > \in i(t)(c),\<l , ( \ell , u ] \ > \in i(t)(c ) \}$ ] + and @xmath144 be the greatest of the set + @xmath145 . if there is any interpretation @xmath2 in @xmath5 where there is not some @xmath108 s.t .
@xmath146 then add @xmath147 \>$ ] to @xmath148 .
if @xmath149 and @xmath150 then add @xmath151 \>$ ] to @xmath148 .
if @xmath149 and @xmath152 then add @xmath153 to @xmath148 .
if @xmath154 and @xmath152 then add @xmath155 to @xmath148 . finally , if @xmath154 and @xmath150 then add @xmath156 \>$ ] to @xmath148
clearly , @xmath157 . in the next part of the proof ,
we show we can create @xmath158 as follows .
we build interpretation @xmath85 . for
each @xmath137 let @xmath138 be the greatest of the set + @xmath139 \
\in i(t)(c),\<l , [ \ell , u ) \ > \in
i(t)(c ) \}$ ] and @xmath140 be the greatest of the set + @xmath141 \ >
\in i(t)(c),\<l , ( \ell , u ) \ > \in i(t)(c ) \}$ ] . then , for each @xmath137 let @xmath142 be the least element of the set @xmath143 \ > \in i(t)(c),\<l , ( \ell , u ] \ > \in i(t)(c ) \}$ ] and @xmath144 be the least of the set @xmath145 . if @xmath159 or @xmath160 then add @xmath161 to @xmath148 . if @xmath154 and @xmath162 then add @xmath151 \>$ ] to @xmath148
. if @xmath154 and @xmath163 then add @xmath153 to @xmath148 . if @xmath149 and @xmath163 then add @xmath155 to @xmath148 . finally , if @xmath149 and @xmath162 then add @xmath156 \>$ ] to @xmath148
clearly , @xmath164 . + as both @xmath136 and @xmath158 exist and are clearly in @xmath5 then the statement follows .
[ [ a - single - application - of - gamma - can - be - computed - in - polynomial - time ] ] a single application of @xmath111 can be computed in polynomial time ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ for interpretation @xmath2 , @xmath165 can be computed by conducting @xmath166 satisfaction checks where @xmath118 is the maximum in - degree of a node in the network .
( this combined with the assumption that the influence function is computed in constant time results in polynomial time computation for a single application of @xmath111 . )
we note that a given rule will require the most satisfaction checks , as a rule will potentially affect a network atom of a certain label for each vertex - time point pair . by the definition of @xmath167
, a given rule clearly causes no more than @xmath168 satisfaction checks .
as the number of rules is no more than @xmath169 , the statement follows . for a given vertex @xmath119
, we will use the notation @xmath120 to denote the number of incoming neighbors ( of any edge type ) and @xmath170 .
first we show that for a given @xmath121 and @xmath34 , a given rule @xmath122 can tighten the bound on a network atom formed with @xmath123 no more than @xmath124 times .
this is because a given rule adjusts the bound on a network atom based on the number of eligible and qualifying neighbors , which are bounded by @xmath171 respectively . at each application of @xmath111
, at least one network atom must tighten .
hence , as there are only @xmath172@xmath125 tightenings possible , this is also the bound on the number of applications of @xmath111 .
suppose , bwoc , that @xmath173 . then , there exists some @xmath34 , @xmath97 and @xmath20 s.t .
@xmath174 , @xmath175 , and @xmath176 s.t .
@xmath177 and @xmath95 .
there are four things that affect @xmath178 : facts , rules , integrity constraints and @xmath179 . clearly , we need not consider the effect that either facts or @xmath179 have on @xmath178 , as @xmath2 satisfies all facts and @xmath128
. we also note that a given integrity constraint imposed by definition [ ibound ] can tighten @xmath178 no more than the associated bound in any model .
hence , there must be some rule @xmath180 that causes @xmath178 to become less than @xmath108 .
as @xmath181 , we know that @xmath182 . as a result , we have @xmath183 and @xmath184 .
further , as @xmath185 and no rule can modify a non - fluent atom , we have claim : if program @xmath33 is consistent then @xmath126 is a model of @xmath33 .
+ suppose , bwoc , that there is a fact in @xmath33 that @xmath126 does not satisfy .
however , by the definition of @xmath111 and the definition of a fact , @xmath126 must satisfy all facts as the bound on the weight associated with each fact is included in the intersection .
further , we can also see by the definition of @xmath111 that @xmath126 strictly satisfies all non - fluent facts in @xmath33 .
we also note that the final application of the @xmath111 operator ensures that all integrity constraints are satisfied by @xmath126 . now ,
suppose , bwoc , that there is a rule in @xmath33 that @xmath126 does not satisfy .
however , with each application of @xmath111 , for each rule , we include the bound on the weight returned by the @xmath193 function for each time step in the target time step associated with that rule . as @xmath111 is applied to convergence , and new bounds are intersected with each application , then we know that all time points in any associated target time set are considered in the intersection .
+ proof of theorem : the above claim tells us that @xmath194 . now consider interpretation @xmath2 s.t .
as @xmath195 , multiple applications of lemma [ boundlemma ] tell us that @xmath196 .
hence , the statement follows .
claim 1 : if @xmath33 is consistent , then @xmath131 is a canonical model of @xmath33 .
+ clearly , @xmath197 satisfies all facts and integrity constraints in @xmath33 .
hence , we shall consider programs that only consist of rules in this proof .
we say @xmath2 @xmath123-canonically satisfies @xmath33 iff @xmath2 canonically satisfies @xmath198 .
clearly , @xmath2 canonically satisfies @xmath33 if for all @xmath199 , @xmath33 @xmath123-canonically satisfies by @xmath2 .
we say that @xmath2 is an @xmath200-canonically consistent interpretation if for @xmath20 , for the first @xmath201 , @xmath202 where @xmath203 .
consider some @xmath204 and @xmath20 .
clearly , @xmath2 is an @xmath205-model for @xmath33 .
let us assume , for some value @xmath206 , that @xmath2 is an @xmath207 model for @xmath33 .
let time point @xmath208 be the @xmath206-th element of @xmath209 .
consider the time step before time @xmath208 is considered in the for - loop at line [ bigfor ] of , which causes the condition at line [ condline ] to be true . by line [ newcurfree ] , @xmath210[l]$ ] .
this means that @xmath208 is a member of both . hence ,
when @xmath208 is considered at line [ bigfor ] , the condition at line [ condline ] is true , causing @xmath211 and as the element @xmath53 is not changed here , we have shown the @xmath2 is an @xmath200-model for @xmath33 . by the for - loop at line [ vlloop ] , for all @xmath199 and @xmath212
, @xmath2 is an @xmath200-model for @xmath33 .
hence , at the for - loop at line [ bigfor ] , we can be assured that for @xmath199 and @xmath212 that @xmath2 @xmath213 satisfies @xmath33 which means that @xmath2 canonically satisfies @xmath33 + claim 2 : if @xmath2 is a canonical model for @xmath33 , + @xmath214 is an interpretation that also strictly satisfies all non - fluent facts in @xmath33 , and @xmath215 is @xmath216 after being manipulated in lines [ insideforbegin]-[insideforend ] of , then @xmath217 .
we note that by the definition of satisfaction of a non - fluent fact , and the fact that both @xmath216 and @xmath2 must strictly satisfy all non - fluent facts in @xmath33 , we know that for all @xmath20 and @xmath34 that : @xmath218 let us assume that lines [ insideforbegin]-[insideforend ] of the algorithm are changing @xmath216 when the outer loop is considering time @xmath208 and that the condition at line [ condline ] is true . clearly , @xmath219 as a result , for any @xmath220 pair considered at this point by the algorithm , if @xmath221 and @xmath222 then we have @xmath223 . by the algorithm ,
if we have @xmath224 and @xmath225 we have that @xmath226 . as + @xmath227 , we know that @xmath228 . as a result , we have @xmath217 , completing the claim .
proof of theorem : as initially @xmath229 and @xmath196 by theorem [ minmodelgamma ] , we note that the algorithm changes + @xmath216 either by applying @xmath111 or manipulating it in lines [ insideforbegin]-[insideforend ] , which tells us ( by claim 2 ) that for all models @xmath2 of @xmath33 that + @xmath230 . since by claim 1 we know that + @xmath231 , the statement of the theorem follows . one way in which can be used is looking at proliferation of research on different topics .
we look at research conducted on niacin , an organic compound commonly used for increasing levels of high density lipoproteins ( hdl ) . using thomson reuters web of knowledge ( http://wokinfo.com )
we were able to extract information on @xmath232 articles about niacin .
this information was then processed using the science of science ( sci@xmath233 ) tool ( http://sci2.cns.iu.edu ) to extract numerous different networks such as author by paper networks , citation networks , and paper by subject networks .
each paper has attributes about when it was published , what journal it was published in , and what subjects the paper was about . during the first time period there is a total of @xmath234 papers with @xmath235 different authors and @xmath236 connections based on an author being cited as an author of a given paper . during the second time period , there is a total of @xmath132 papers with @xmath133 different authors and @xmath237 connections . | the modeling of cascade processes in multi - agent systems in the form of complex networks has in recent years become an important topic of study due to its many applications : the adoption of commercial products , spread of disease , the diffusion of an idea , etc . in this paper , we begin by identifying a desiderata of seven properties that a framework for modeling such processes should satisfy : the ability to represent attributes of both nodes and edges , an explicit representation of time , the ability to represent non - markovian temporal relationships , representation of uncertain information , the ability to represent competing cascades , allowance of non - monotonic diffusion , and computational tractability .
we then present the @xmath0 language , a formalism based on logic programming that satisfies all these desiderata , and focus on algorithms for finding minimal models ( from which the outcome of cascades can be obtained ) as well as how this formalism can be applied in real world scenarios .
we are not aware of any other formalism in the literature that meets all of the above requirements .
[ representation languages ] |
the reheating period in the early universe stands out as a challenge to theorists due to the close interrelationship of nonlinear and gravitational phenomena in its unfolding ( see @xcite , henceforth
papers 1 - 5 ) .
the observation that , due to the high occupation numbers produced during preheating , most of the physics of reheating may be understood in terms of the behavior of nonlinear classical waves@xcite has been the key to sustantial progress .
the authors of papers 1 - 5 have undertaken systematic numerical simulations of the behavior of matter fields during reheating , finding that the reheating period is actually composed of three consecutive phases : an early one or preheating , where the dominant effect is the parametric amplification of matter fields out of the dynamical inflaton and gravitational backgrounds @xcite , an intermediate stage where the dominant phenomenon is the redistribution of energy among matter field modes through rescattering ( in the terminology of papers 1 - 5 ) , and a final stage where thermal equilibrium sets in . during the intermediate stage ,
spectra of occupation numbers for the matter fields reduce to simple power laws both in the infrared and ultraviolet limits .
as noted in papers 1 - 5 , this behavior suggests a connection between the physics of reheating and the phenomena of weak turbulence @xcite , but to the best of our knowledge no theoretical prediction for the exponents involved is available .
our goal is to provide these theoretical estimates . in this paper
we shall follow this same trend of ideas , by observing that , from the macroscopic point of view , a stochastic ensemble of classical waves may be described by a conserved energy momentum tensor subject to the second law of thermodynamics .
there is therefore an equivalent fluid description , consisting of a fluid whose energy momentum tensor and equation of state reproduce the observed ones for the microscopic fluctuations . solving the dynamics of this equivalent fluid yields answers to all relevant questions concerning the behavior of the actual fluctuations .
an immediate consequence of energy momentum conservation and the second law is that when velocities are low , the phenomenological fluid may be described within the eckart theory of dissipative fluids @xcite ( for an analysis of the limitations of eckart s theory see @xcite ) .
it follows that it obeys a continuity equation and a curved space time navier - stokes one .
turbulent spectra found in numerical simulations correspond to the self - similar solutions discussed long ago by chandrasekhar @xcite .
they appear in the discussion of decaying turbulence ( which is the case relevant to cosmology ) , as opposed to turbulence driven by some external means .
the chandrasekhar solutions are built on the heisenberg closure hypothesis @xcite ( see @xcite for a general discussion of turbulence ) .
they were generalized to friedmann - robertson - walker ( frw ) backgrounds by tomita _
et al _ @xcite .
these solutions agree with the kolmogorov 1941 theory in the inertial range @xcite , failing to reproduce observations for very small eddies .
fortunately , we are most interested in the opposite limit of very large eddies , where it is trustworthy ( we wish to point out that the applicability of kolmogorov s spectrum to large scale turbulence should not be taken for granted @xcite ) . with minor adjustments ,
tomita s analysis of turbulence decay in frw space times also provides a solution to the evolution of our equivalent fluid .
the rest of the paper is organized as follows . in next section
we provide a brief summary of hydrodynamics in flat and expanding universes , in order to set up the language for the rest of the paper , and introduce the self similar solutions .
in section iii we proceed to discuss the equivalent fluid description of field fluctuations , and how to extract the particle spectrum therefrom . in section
iv we place the self - similar solutions in the context of reheating .
we state our main conclusions in the final section .
we provide a rough estimate of the shear viscosity during reheating in the appendix .
the equations governing the dynamics of a fluid in local thermodynamic equilibrium are the continuity and navier - stokes ones , which , in the case of flat space time , read : @xmath3 @xmath4 where we have assumed incompressibility , valid when typical velocities are much smaller than the sound velocity ; @xmath5 is the kinematic shear viscosity .
the transition from laminar to turbulent motion can be universally described by the dimensionless reynolds
number : @xmath6 where @xmath7 is a typical velocity and @xmath8 a typical length scale .
this number represents the order of magnitude of the ratio of the inertial to the viscous term .
low reynolds numbers correspond to laminar motion , while high ones suggest turbulent behavior .
in general , the velocity profile displays variations in space and time .
this implies that the flow must be described probabilistically .
thus , each quantity involved in ( [ contpl]-[nspl ] ) is divided in its mean value and a fluctuation from it ; for example , we write @xmath9 , where @xmath10 stands for the fluctuating part of the velocity . in the case where motion is isotropic ,
the mean value @xmath11 for the velocity must be zero , since otherwise there would be a preferred direction . to analyze the system s behavior
, we define the two - point one - time correlation function for the velocity : @xmath12 in the case of homogeneous and isotropic motion , this correlation must be only a function of the time @xmath13 and the distance between @xmath14 and @xmath15 , i.e. @xmath16 , where @xmath17 observe that @xmath18 ( summation over repeated indices must be understood ) is twice the average energy density of the flow at time @xmath13 . from ( [ nspl ] )
we obtain the equation that this correlation must obey , namely : @xmath19 where @xmath20 and @xmath21 the tensor @xmath22 comes form the inertia term in navier - stokes equation and , as it involves a product of third order in the velocity , reflects the fact that there is not a close set of equation for the correlations of successive orders but there is a hierarchy of equations instead .
the problem of closing that hierarchy is known as the moment closure problem @xcite .
let us call @xmath23 the fourier transform of @xmath24 .
then the energy density becomes @xmath25 where @xmath26 is the energy density stored in eddies of size @xmath27 defining @xmath28 as the fourier transform of @xmath22 , we obtain from ( [ ecrij ] ) the equation of balance of the energy spectrum : @xmath29 where @xmath30 the inertia term @xmath31 is the one that contains the mode - mode interaction , and its effect is to drain energy form the more energetic modes -typically the bigger ones- to the ones where there is major viscous dissipation -the smaller ones- . for a curved space - time , in particular a friedmann - robertson - walker ( frw ) universe with zero spatial curvature ( @xmath32 ) , the generalization of the above arguments has been considered by many authors @xcite .
we follow tomita _ et al .
_ s analysis @xcite , in which they obtain the solution for the energy spectrum in the case of homogeneous , isotropic and incompressible turbulence . in a generic space time , we describe fluid flow from the energy density @xmath33 , pressure @xmath34 and four velocity @xmath7 .
the symmetries of the frw solution suggest using instead the commoving three velocity @xmath35 if @xmath36 the flow is non relativistic , and if @xmath37 it is incompressible ( @xmath38 ) .
later on , we shall also use the physical three velocity @xmath39 the corresponding continuity and navier - stokes equations for a robertson - walker background are obtained by the condition of conservation of the energy - momentum tensor @xcite . for a non relativistic incompressible fluid , with shear viscosity @xmath40 ( but no bulk viscosity ) , these reduce to : @xmath41 @xmath42 { \bf u=-}\frac{\nabla p}{a^2\left ( p+\rho \right ) } + \frac 1{a^2}\nu \nabla ^2 { \bf u } \label{nscur}\ ] ] where we have assumed that @xmath43 depends only on time . for the physical three velocity @xmath44 ,
the corresponding navier - stokes equation reads : @xmath45 { \bf v=-}\frac{\nabla p}{a\left ( p+\rho \right ) } + \frac 1{a^2}\nu \nabla ^2{\bf v } \label{nscurv}\ ] ] in obtaining ( [ contcur])-([nscurv ] ) we have neglected possible perturbations to the frw metric .
the corresponding equations considering fluctuations in the metric ( @xmath46 ) have been obtained by weinberg @xcite .
the continuity equation is not corrected by gravitational perturbations , while in the navier - stokes equation the metric fluctuations appear explicitly only within the shear viscosity term .
it can be demonstrated that these terms involving metric fluctuations are negligible for scales that are inside the horizon @xcite . for scales bigger than the hubble radius ,
since dissipation through viscosity is not effective anyway , we may still use the unperturbed navier - stokes equation .
the operation of fourier transforming in the case of a robertson - walker cosmology is done in terms of commoving wave - numbers . in doing so
, the following equation for the energy spectrum is obtained : @xmath47 where the relationship between @xmath48 and @xmath23 as well as between @xmath31 and @xmath49 is the same as that for a flat space time , if we define @xmath50 and @xmath22 from correlations of physical quantities , as follows : @xmath51 as we have seen in the previous section , the key element in the description of the flow is the energy spectrum @xmath52 ( eq.([et ] ) ) , which is the solution of the balance equation ( eq.([ecesp ] ) ) . in it ,
the right hand side contains the viscous dissipation as well as the inertial force @xmath53 .
the overall effect of this term is to transfer energy from a given scale to smaller ones through mode - mode coupling ; thus it is natural to model the action of the inertia term as a source of viscous dissipation , where the effective turbulent viscosity for a given mode depends on the motion of all smaller eddies @xcite . by providing closure ,
that is , writing this effective viscosity in terms of the spectrum itself , a closed evolution equation for @xmath52 is obtained . concretely ,
heisenberg @xcite proposed the ansatz @xmath54 where @xmath55 and @xmath56 is a dimensionless constant . with this hypothesis ( known as the heisenberg hypothesis ) as the solution to the closure problem ,
chandrasekhar @xcite has obtained the energy spectrum for decaying turbulence , assuming that there is a stage in the decay where the bigger eddies have sufficient amount of energy to maintain an equilibrium distribution , thus requiring that the solution for the spectrum should be self similar . with this consideration into account he obtained an energy spectrum : @xmath57 where @xmath58 and @xmath59 are initial conditions ( namely , the wave number corresponding to the bigger eddy and its typical time of evolution ) .
the function @xmath60 obeys the equation @xmath61 dx^{\prime } = \left\ { \nu k_0 ^ 2t_0+\int_x^\infty \frac{% \sqrt{f(x^{\prime } ) } } { x^{\prime \ 3/2}}dx^{\prime } \right\ } \int_0^xf(x^{\prime } ) x^{\prime \
2}dx^{\prime } \label{ecf}\ ] ] which predicts a kolmogorov type behavior for an inviscid fluid ( @xmath62 , @xmath63 ) in the ultraviolet limit : @xmath64 while for nonzero viscosity : @xmath65 in the infrared limit , @xmath60 has the universal behavior @xmath66 @xmath67 , and thus we find a linear energy spectrum for @xmath68 chandrasekhar s self similar solutions are easily generalized to flows in expanding universes .
the dependence on time and wave - number for the self similar energy spectrum is @xcite @xmath69 where @xmath70 and @xmath71 are respectively the taylor s microscale and an average turbulent velocity , defined as : @xmath72 the second equation implies the normalization condition @xmath73 to obtain a self similar flow reducing to eq .
( [ flatspacelim ] ) in the flat space limit , we must require the time evolution laws : @xmath74 the equation which determines de function @xmath75 in ( [ speccur1 ] ) turns out to be @xmath76 dx^{\prime } = \left\ { \frac 25+a\int_x^\infty \frac{\sqrt{f(x^{\prime } ) } } { x^{\prime \ 3/2}}dx^{\prime } \right\ } \int_0^xf(x^{\prime } ) x^{\prime \
2}dx^{\prime } \label{curvedchandra}\ ] ] where @xmath77 is a constant .
this equation has the same structure as in flat space time , eq . ( [ ecf ] ) , which means that assuming heisenberg s hypothesis the spectrum is linear in @xmath78 for length scales much bigger than the taylor s microscale .
observe that for flat space time , the proportionality between the integral up to a certain wave number @xmath78 of the inertia and the viscous forces is given by ( [ heis1 ] ) and ( [ heis2 ] ) .
in the case of a frw space time , the autosimilar solution required by tomita _
( [ speccur1 ] ) needs a time dependent @xmath79 in eq .
( [ heis1 ] ) , defined by the analog of ( [ heis2 ] ) with @xmath56 replaced by @xmath80 .
let us analyze in more detail the solutions of eq .
( [ curvedchandra ] ) .
we assume the normalization eq .
( [ normalization ] ) . by taking the @xmath81 limit in eq .
( [ curvedchandra ] ) we find @xmath82 taking a derivative of eq .
( [ curvedchandra ] ) we get @xmath83 where @xmath84 @xmath85 let us consider first the @xmath86 limit .
assume @xmath87 . the left hand side of eq .
( [ differential ] ) tends to a finite limit @xmath88 . on the right hand side , @xmath89 and @xmath90
behave as @xmath91 , so if @xmath92 this side goes to zero .
we must therefore have @xmath93 , and @xmath94 were @xmath95 is some constant . in the @xmath96 limit ,
assume again a power law behavior @xmath97 .
now @xmath98 , so we must have @xmath99 from eq . ( [ normalization2 ] ) we know that in this limit @xmath100 , so we must have @xmath101 and @xmath102 taking into account both limiting behaviors and eq . ( [ normalization ] ) ,
the function @xmath60 may be approximated as @xmath103 = \frac{x}{\left [ \alpha + \beta x^{4}\right ] ^{2}},\qquad \alpha = \left [ \frac{25a\pi ^{2}}{128}\right ] ^{1/3}\qquad \beta = \frac{2 } { 25a } \label{fapprox}\ ] ]
after establishing the basic necessary notions for the description of hydrodynamic flows , our goal is to associate an equivalent fluid description to field fluctuations , and to derive the particle spectrum therefrom .
our first step is to obtain the energy density , pressure and velocity of this fluid as functionals of the quantum state of the field . for simplicity
, we shall consider the theory of a single , self - interacting scalar field @xmath104 , minimally coupled to gravity .
the action is @xmath105 \right\}\ ] ] where @xmath106 $ ] ia renormalized effective potential .
the energy - momentum tensor is associated to the heisenberg operator @xmath107 \right\}\ ] ] the macroscopic behavior of the field , however , may be described in terms of a c - number energy - momentum tensor @xmath108 where @xmath109 and @xmath110 is a stochastic component with zero mean and self - correlation @xmath111 in these equations , @xmath112 denotes a stochastic average , while @xmath113 is the average with respect to the quantum state of the field .
following landau , we define the fluid four - velocity @xmath114 and energy density @xmath115 as the ( only ) time - like eigenvector of @xmath116 and ( minus ) its corresponding eigenvalue @xmath117 introducing the pressure @xmath118 as given by the equilibrium equation of state ( our theory does not have a conserved particle number current , and therefore the equation of state is barotropic ) , we may decompose @xmath119 where @xmath120 and by construction @xmath121 .
since @xmath122 vanishes by definition in the equilibrium state , it may be parametrized in terms of deviations from equilibrium .
remaining within the so - called _ first order formalism _
@xcite , we may write @xmath123 where @xmath124 \label{hmunu}\ ] ] and @xmath125 and @xmath126 are the _ shear _ and _ bulk _ viscosity coefficients , respectively .
let us decompose each quantity in a mean component ( denoted by a @xmath95 subscript ) and a fluctuation ( denoted by a @xmath127 ) .
if the quantum state shares the symmetries of the frw background , then @xmath128 . since @xmath129 holds identically ( as oppossed to in the mean ) we must have @xmath130 @xmath131 = 0\ ] ] the second equation shows that @xmath132 is a higher order fluctuation with respect to @xmath133 . if we remain at linear order , then , we may approximate @xmath134 observing that all mean values are homogeneous and isotropic , we see that @xmath135 is also a higher order fluctuation .
we find @xmath136 and therefore the velocity correlation @xmath137 this is the key equation linking the quantum and stochastic descriptions . to estimate the velocity correlation ,
let us assume that , after integrating out the hard modes , the soft modes of interest may be described in terms of quasi free , long lived excitations with an effective mass @xmath138 then @xmath139 the fluctuations are gaussian to a very good approximation , and therefore @xmath140 where @xmath141 is the positive frequency propagator @xmath142 ( @xmath143 stand for derivatives with respect to the first argument , while @xmath144 stand for derivatives with respect to the second argument of @xmath145 ) .
let us decompose the soft field into modes @xmath146 at any time @xmath147 we may introduce positive frequency adiabatic modes defined by @xcite @xmath148 where @xmath149 and decompose the mode amplitude @xmath150 into positive and negative frequency components @xmath151 @xmath152 let us define the spectrum @xmath153 ( because the quantum state is isotropic , the spectrum depends only on @xmath78 ) and assume that @xmath154 ( this happens , for example , if the different modes acquire random phases through interaction with an environment ) . then @xmath155 and @xmath156 observe that only the vacuum part contributes in the coincidence limit @xmath157 .
in the large occupation numbers regime we are interested in , this is negligible , and we get @xmath158 where @xmath159 @xmath160 we may now write down the fourier transform of the velocity self correlation @xmath161 in principle , this is an integral equation relating the spectrum to the energy self - correlation .
it may be simplified as follows . for small @xmath162 ,
the integral reads @xmath163 while for large @xmath34 we get @xmath164 since the spectra we are considering fall off much faster than @xmath165 ( or even @xmath166 ) at large @xmath167 we may estimate that the contribution from eq .
( [ smallp ] ) is larger than eq .
( [ largep ] ) ; moreover , extending the integral to the whole momentum space , @xmath168 and we get @xmath169 finally , we may relate the particle spectrum to the turbulent energy spectrum @xmath170
in the previous sections , we have analyzed on one hand self similar turbulent flows in expanding universes , and on the other hand have given the rule to translate the turbulent energy spectrum into a particle number distribution .
we must now show that the foregoing analysis is relevant to plausible models of the reheating period , and use it to predict the likely shape of the final particle spectrum . at this point
, it is convenient to be more precise about the model of reheating we have in mind ( for further details , see papers 1 - 5 ) .
we assume that inflation is driven by an inflaton field @xmath171 with an effective potential which may be parameterized as @xmath172 the self coupling @xmath173and at the end of inflation @xmath174 ( in natural units ) .
this is the dominant contribution to the energy density , so it fixes the scale for the hubble constant @xmath90 : @xmath175 . during preheating , a large fraction ( nearly all ) of this energy
is transferred to the matter fields .
these are described by an effective degree of freedom @xmath104 which self interacts with an effective dimensionless coupling constant @xmath176 for simplicity , we shall model this self interaction as a @xmath177 theory . at times
@xmath147 at the end of preheating , matter excitations are distributed with a smooth spectrum @xmath178 , where @xmath78 is a comoving wave number and the characteristic momentum @xmath179 ( meaning that the relevant modes are inside the horizon ) .
we wish to find the form of the spectrum as follows from the assumption that its further evolution will be self - similar .
let us estimate the energy density in the matter fields at @xmath147 as @xmath180 .
the self interaction of the matter fields induces a mass @xmath181 .
we shall assume @xmath182 in which case @xmath183 we shall assume that this is the dominant contribution to the matter self - energy .
the condition @xmath184 is equivalent to @xmath185 ( this equation will be important below ) .
observe that @xmath186 since @xmath187 is already less than one , this is not unduly restrictive on @xmath188 ( see papers 1 - 5)@xmath189 also observe that @xmath190 redshifts with the cosmological evolution , so these estimates are not affected by hubble flow .
integrating equation ( [ fluid2quantum ] ) over @xmath78 , we find the mean velocity in the equivalent fluid flow as @xmath191 where @xmath192 ( @xmath193 ) is the speed of sound , so the flow may be regarded as incompressible .
the shear viscosity of the matter fields may be estimated as ( see appendix ) @xcite @xmath194 leading to a kinematic viscosity @xmath195 with a typical velocity of order one , and a typical lenght of order @xmath196 we get the reynolds number @xmath197 ( cfr .
( [ lambdan ] ) ) . of course
, we do not have a microscopic justification for heisenberg s closure , so we shall take @xmath77 in eq .
( [ curvedchandra ] ) as a free parameter .
as it turns out , agreement with the numerical results reported by papers 1 - 5 is obtained for @xmath198 which , as expected , corresponds to reynolds numbers of order of one .
it s worth to point out that chandrasekhar s solution with the function @xmath60 obeying eq .
( [ curvedchandra ] ) is an exact solution that relies only on heisenberg s closure hypotesis eq .
( [ heis1 ] ) .
so if we trust this hypotesis , we can get the spectrum by solving eq .
( [ curvedchandra ] ) for any reynolds number .
the distribution of occupation numbers is found from eq .
( [ fluid2quantum ] ) , where the energy spectrum is given by eq .
( [ speccur1 ] ) . in the light field limit @xmath199
we find @xmath200 .
( [ fsmall ] ) implies @xmath201 for @xmath202 and eq .
( [ flarge ] ) implies @xmath203 for @xmath204 .
this theoretical prediction for the exponents involved is the main result of this paper . in fig .
1 we show the full particle spectrum based on the approximation eq .
( [ fapprox ] ) for the function @xmath60 . we have scaled the plot to make it easiest to compare with papers 1 - 5 .
momentum is measured in units of @xmath205 ( where @xmath206 is the scale factor ) @xcite , and we have chosen @xmath207 observing that in the ultrarelativistic limit the speed of flow @xmath208 and @xmath209 , integrating the particle density times @xmath210 shows that the total energy density in the flow is @xmath211 .
the equivalent black - body temperature is then somewhat less than @xmath212 , which is a reasonable value for the reheating era , and high enough to justify the neglect of all masses .
the analysis leading to the chandrasekhar solutions begins from the navier - stokes equations for the fluid , which are equivalent in this context to energy - momentum conservation .
therefore our model requires that it be possible to assign to the fluid an independently conserved energy - momentum tensor .
this is not exactly the same as requiring that the homogeneous mode has totally decayed , but it does mean that there is no significative energy transfer from the homogeneous mode to the fluid , either through parametric amplification or otherwise .
the plots of total particle density and effective masses presented in ref .
@xcite suggest that this condition is met early , then follows an intermediate stage dominated by rescattering , and finally the thermalization stage .
comparison with the results in papers 1 - 5 is meaningful only in the intermediate phase .
decaying turbulence is necessarily a transient phenomenon .
as time evolves , we expect the field will eventually thermalize , and the spectrum will get closer to a rayleigh - jeans tail , @xmath213 when masses are negligible .
the several plots presented in ref .
@xcite , where indexes go from @xmath214 to slightly over @xmath215 , capture the transition from turbulence to equilibrium . since the same plots show that earlier spectra are steeper ( see also fig . 1 in ref .
@xcite ) this is in satisfactory agreement with the prediction from self - similar flows .
fig . 1 : log - log plot of the particle spectrum , as given by eq .
( [ fluid2quantum ] ) .
the energy spectrum is given by ( [ speccur1 ] ) , where the function @xmath60 is given by eq .
( [ fapprox ] ) .
we have chosen @xmath216 , the taylor microscale @xmath217 , and have scaled the spectrum to make it easiest to compare against the results presented in @xcite and @xcite .
in this paper , we have shown that the self similar flows studied by heisenberg , chandrasekhar and tomita may be used to provide an interpretation of the turbulent spectra found in papers 1 - 5 .
the hydrodynamic model predicts scale invariant spectra @xmath218 both in the infrared and ultraviolet limits , with @xmath219 in the former , and @xmath220 in the latter regime .
agreement with the early time results presented in papers 1 - 5 is satisfactory .
the connection of hydrodynamics to the behavior of fluctuations during reheating has interest of its own , as it provides an alternative to brute force quantum field theoretic calculations , and also yields physical insight on the macroscopic behavior of quantum fields in the early universe .
the equivalent fluid method may be used to advantage also in other regimes , such as the inflationary period itself @xcite . moreover
, it opens up a wealth of new phenomena , such as intermittence @xcite and shocks @xcite , which are not apparent in the customary treatments .
we will continue our research in this field , which promises a most rewarding dialogue between cosmology , astrophysics , and nonlinear physics at large .
we thank mario castagnino for rekindling our interest in the subject .
ec has been partially supported by universidad de buenos aires , conicet , anpcyt under project pict-99 03 - 05229 and fundacin antorchas .
mg has been partially supported by national science foundation grant phy97 - 22022 .
in order to estimate the viscosity for the matter fields we assume that the correspondence eq .
( [ fluid2quantum ] ) between the fluid and particle spectra allows us to associate a solution of the hydrodynamical evolution equations for the former to a self - similar solution of the kinetic equation for the latter ( cfr .
@xcite ) . we may then estimate the transport coefficients by adopting the same methods usually applied in equilibrium @xcite .
these are discussed in detail , in the quantum field theory context , in references @xcite .
@xmath224 & \sim & g^{2}\int dk_{2}dk_{3}dk_{4}\;\delta \left ( k_{1}+k_{2}-k_{3}-k_{4}\right ) \nonumber \\ & & \left\ { \left ( 1+n_{k1}\right ) \left ( 1+n_{k2}\right ) n_{k3}n_{k4}-\left ( 1+n_{k3}\right ) \left ( 1+n_{k4}\right ) n_{k1}n_{k2}\right\}\end{aligned}\ ] ] there is a vast literature on preheating .
some representative works are l. kofman , a. linde and a. starobinsky , phys .
lett * 73 , * 3195 ( 1994 ) ; y. shtanov , j. traschen and r. brandenberger , phys . rev .
* d51 , * 5438 ( 1995 ) ; d. boyanovsky , h. de vega , r. holman and j. salgado , phys . rev .
* d54 , * 7570 ( 1996 ) ; s. ramsey and b. l. hu , phys . rev . *
d56 , * 678 ( 1997 ) ; p. greene , l. kovman , a. linde and a. starobinsky , phys . rev .
* d56 , * 6175 ( 1997 ) ; b. bassett , d. kaiser and r. maartens , phys . lett . *
b455 * , 84 ( 1999 ) ; b. bassett , c. gordon , r. maartens and d. kaiser , phys . rev .
* d61 , * 061302 ( 2000 ) ; f. finelli and s. khlebnikov , e - print hep - ph/0107143 .
s. jeon , phys . rev .
* d47 * , 4586 ( 1993 ) ; s. jeon , phys . rev . *
d52 * , 3591 ( 1995 ) ; s.jeon and l.g.yaffe , phys . rev . * d * * 53 * , 5799 ( 1996 ) ; e. calzetta , b. l. hu and s. ramsey , phys . rev . *
d61 , * 125013 ( 2000 ) .
s. chapman and t. cowling , _ the mathematical theory of non - uniform gases _ ( cambridge university press , london , 1990 ) ; r. liboff , _ kinetic theory _
( john wiley , new york , 1998 ) ; k. huang , _ statistical mechanics _ ( john wiley , new york , 1987 ) . | we show that the turbulent particle spectra found in numerical simulations of the behavior of matter fields during reheating admit a simple interpretation in terms of hydrodynamic models of the reheating period .
we predict a particle number spectrum @xmath0 with @xmath1 for @xmath2 principal pacs no 98.80.cq ; additional pacs nos : 04.62.+v , 11.10.wx , 47.27 |
recent developments in spectroscopy have led to very precise experimental values for the @xmath1 lamb shift and the rydberg constant @xcite , so that now the lamb shift provides the best test of quantum electrodynamics for an atom .
these achievements have spurred great theoretical efforts aimed at matching the current experimental accuracy ( for a review of the present status and recent developments in the theory of light hydrogenic atoms , see @xcite ) .
the theoretical prediction is expressed in terms of three small parameters : @xmath2 describing effects due to the binding of an electron to a nucleus of atomic number @xmath3 ; @xmath4 ( frequently accompanied by @xmath5 ) from electron selfinteractions ; and the ratio of electron to nucleus masses .
the lamb shift is of the order @xmath6 ; all corrections through the second order in the small parameters are known , as well as some of the third order @xcite .
another source of corrections is the spatial distribution of the nuclear charge . even for hydrogen , the experimental uncertainty in the measurement of the proton root mean
square charge radius poses an obstacle for further theoretical progress .
fortunately , measurements can be performed also with the muonic hydrogen whose spectrum is much more sensitive to the proton radius .
a comparison of the theoretical prediction @xcite and anticipated new measurements @xcite is expected to soon improve the knowledge of this crucial parameter .
in this paper we focus on the second - order radiative - nonrecoil contributions to the lamb shift of order @xmath0 . the total result for the corrections of this order
was presented first in @xcite and improved in @xcite .
our full result is compatible with the previous ones and has better precision . when comparing contributions from individual diagrams with @xcite , however , we find small discrepancies in some cases .
in section [ eval ] we present the details of our approach , and in section [ res ] we present our results . in appendix [ app ]
we show the results for the master integrals , and in appendix [ appb ] new analytic results for two diagrams .
we consider an electron of mass @xmath7 orbiting a nucleus of mass @xmath8 and atomic number @xmath3 , where @xmath3 is assumed to be of such a size that @xmath9 is a reasonable expansion parameter .
we are interested in corrections to the lamb shift of order @xmath0 and leading order in @xmath10 , given by @xmath11 where @xmath12 is the squared modulus of the wave function of an @xmath13 bound state with principal quantum number @xmath14 ( @xmath15 is the reduced mass of the system ) , and @xmath16 is the momentum space representation of the amplitude of the interaction between the electron and the nucleus at orders @xmath17 and @xmath18 . both particles are considered to be at rest and on their mass shell @xcite .
the correction @xmath19 is given by the sum of all the three - loop diagrams presented in figs .
[ vacuum ] and [ other ] . in these figures
, the continuous line represents the electron , and the dashed line represents the interaction with the nucleus .
the reason for this is that for our purposes this interaction can be replaced by an effective propagator . in all diagrams ,
the leading order in @xmath10 comes from the region where all the loop momenta scale like @xmath7 .
the part of the diagrams representing the interaction between the electron and the nucleus at order @xmath20 is given by the sum of the direct and crossed two - photon exchange shown in fig .
[ graphdelta ] . if @xmath21 and @xmath22 are the loop and nucleus momenta , respectively , and @xmath23 , the sum of the nucleus propagators can be approximated at leading order by @xmath24 since the nucleus is considered to be at rest , this gives us a @xmath25 .
together with the propagators of the two photons , this constitutes the effective propagator .
we used dimensional regularization , and renormalized our results using the on - shell renormalization scheme .
for all the photon propagators in figs .
[ vacuum ] and [ other ] we used a general covariant @xmath26 gauge .
the overall cancellation of the dependence on the gauge parameter in the final result provided us with a good check for our calculations . since the gauge invariance of the sum of the subdiagrams in fig .
[ graphdelta ] is trivially and independently fulfilled , for the photonic part of the effective propagator we used the feynman gauge .
the different sets of vacuum polarization diagrams .
each set represents the drawn diagram plus all the possible permutations of its pieces . ] the diagrams involving a two - loop electron self - interaction and vertex corrections . ]
the sum of the direct and crossed diagrams is approximated by an effective propagator ( the double line represents the propagator of the nucleus ) . ] since we are considering free asymptotic states with independent spins , the dirac structures of the electron and the nucleus factorize , and we simplify them by inserting each of the structures between the spinors of the initial and final states and averaging over the spins of the initial states .
we use the program ` qgraf ` @xcite to generate all of the diagrams , and the packages ` q2e ` and ` exp ` @xcite to express them as a series of vertices and propagators that can be read by the ` form ` @xcite package ` matad 3 ` @xcite . finally , `
matad 3 ` is used to represent the diagrams in terms of a set of scalar integrals using custom - made routines . in this way
, we come to represent the amplitude @xmath27 in terms of about 18000 different scalar integrals .
these integrals can be expressed in terms of a few master integrals by means of integration - by - parts ( ibp ) identities @xcite .
using the so - called laporta algorithm @xcite as implemented in the ` mathematica ` package ` fire ` @xcite , we find 32 master integrals .. ] since the program ` fire ` deals only with standard propagators , when using it we worked only with one of the nucleus propagators , instead of the dirac delta of the effective propagator .
that is , instead of working with @xmath28 , we worked with @xmath29 , for example .
working with just one of the propagators is enough for this purpose , as the ibp method is insensitive to the @xmath30 prescription ( remember from eq .
( [ delta ] ) that in our approximation this is the only difference between the two propagators ) . since each diagram in figs .
[ vacuum ] and [ other ] represents the subtraction of two integrals that only differ in the nucleon propagator , when applying the ibp method we can set to zero any resulting integral in which the propagator @xmath29 disappears .
this can be done because the same integral with the propagator @xmath31 instead would give the exact same contribution and thus the difference between the two is zero .
once the reduction to master integrals is complete , we can simply substitute back the delta function in place of the nucleon propagator . in order to calculate the master integrals , we turned the expressions in appendix [ app ] into a representation in terms of feynman parameters .
the procedure we then followed in most cases was to use a mellin - barnes representation @xcite to break up sums of feynman parameters raised to non - integer powers and transform the integrals into integrals of gamma functions over the imaginary axis .
in some cases , we were able to obtain analytical results .
otherwise , we used the ` mathematica ` packages ` mb ` @xcite and ` mbresolve ` @xcite to perform a numerical calculation . for integrals
@xmath32 , @xmath33 , @xmath34 , @xmath35 , @xmath36 , @xmath37 , @xmath38 , and @xmath39 ( cf .
appendix [ app ] ) , the mellin - barnes representation was too cumbersome for a numerical evaluation . in these cases
we used the ` mathematica ` package ` fiesta 1.2.1 ` @xcite with integrators from the ` cuba ` library @xcite to perform numerical computations using sector decomposition @xcite . like ` fire ` , ` fiesta ` can only process standard propagators as input .
this means that we had to use the momentum representation of the integrals with the nucleon propagators instead of the delta function .
we separately calculated the integrals containing @xmath40 and the ones containing @xmath41 instead , and added the two results .
we checked the method by computing with ` fiesta ` some integrals we had already found with ` mb ` and ` mbresolve ` .
the results always agreed .
there was one case , integral @xmath39 , where the ` fiesta ` result for the integral with @xmath41 was numerically unstable .
fortunately , in this case we could find a representation in terms of feynman parameters that we could compute directly using ` cuba ` , without further treatment .
this was possible because the integral is finite , and the representation was free of spurious divergences .
we cross - checked this result using a beta version of ` fiesta 2 ` @xcite , which did not produce the instabilities we encountered in the former version .
we performed an additional cross - check of our results by changing the basis of integrals . to do this
, we took one of the integrals we computed with ` fiesta ` and used the ibp method to express it in terms of a similar integral of our choice ( same as the original one , but with some propagator(s ) raised to different powers ) plus other master integrals we already knew .
we then computed the new integral with ` fiesta ` and checked if the final result for the lamb shift ( or for individual diagrams ) agreed with the calculation in the old basis .
since changing the basis modifies the coefficients of all the integrals involved in the change , the agreement of the results obtained with different bases is a very good cross - check of our calculations .
this cross - check was performed for several integrals . in particular , we changed integrals @xmath42 and @xmath36 , which are the ones limiting our precision , and integrals @xmath35 and @xmath39 . since the last two integrals contain most of the propagators for integral types @xmath43 and @xmath44
, the corresponding changes of basis affect the coefficients of most of the other integrals of the respective type .
our final results for the separate contributions from the vacuum - polarization diagrams of fig . [ vacuum ] and
the diagrams @xmath45@xmath46 of fig .
[ other ] are @xmath47\,,\label{vac}\\ \delta e_{a - s } & = & \frac{{\alpha}^{2}(z{\alpha})^{5}}{\pi n^{3}}{\left(}\frac{\mu}{m}{\right)}^{3}m\,[-7.72381(4)]\,.\label{as}\end{aligned}\ ] ] the best results so far for the vacuum polarization diagrams and for diagrams @xmath45@xmath46 have been published in @xcite ( cf .
@xcite for references of partial results ) and @xcite , respectively .
our results are compatible with them and improve the precision by two orders of magnitude in the case of @xmath48 and a little over one order of magnitude for @xmath49 .
the total result reads @xmath50\,,\label{total}\ ] ] and the corresponding energy shifts for the @xmath1 and the @xmath51 states in hydrogen are @xmath52 .[tvac ] comparison between our results for the different vacuum - polarization sets ( in fried - yennie gauge ) and those of @xcite .
numbers ending in an ellipsis indicate an analytic result , which we show in appendix [ appb ] . [
cols="<,<,<",options="header " , ] choosing the fried - yennie gauge @xcite , we also compared the results from the different sets of vacuum polarization diagrams with those of @xcite , and the results from the individual diagrams @xmath45@xmath46 with those of @xcite .
our results for the vacuum polarization graphs and diagrams @xmath45@xmath46 are presented in tables [ tvac ] and [ tother ] , respectively .
all numbers in the tables are to be multiplied by the prefactor @xmath53 ( note the difference in normalization in @xcite ) .
we found new analytic results for four diagrams .
the results for diagrams @xmath54 and @xmath55 are given in table [ tother ] , while the results for diagrams @xmath56 and @xmath57 , being too lengthy for the table , are presented in appendix [ appb ] .
for completeness , the known analytic results for sets ii and iii of the vacuum polarization diagrams are given in appendix [ appb ] as well .
it should be mentioned that the errors of the results in eqs .
( [ vac ] ) , ( [ as ] ) , and ( [ total ] ) are not obtained from the sum of the errors of the diagrams in tables [ tvac ] and [ tother ] .
once we decompose the problem into the calculation of master integrals , the diagrams are no longer independent , as the same master integral contributes to several different diagrams .
thus , to find the error of our total result , we first sum all diagrams and then sum all the errors of the integrals in quadrature .
we found discrepancies between our results for diagrams @xmath45@xmath46 and those of @xcite .
most of the central values in the second and third column of table [ tother ] lie between @xmath58 and @xmath59 away from each other , but in the case of diagrams @xmath60 and @xmath46 the difference is around @xmath61 , and for diagrams @xmath21 and @xmath62 , it reaches @xmath63 ( we take as @xmath64 the errors of individual diagrams in the third column ) .
we should stress again that our calculation is done using dimensional regularization while the study of ref .
@xcite was performed in four dimensions . even though all the individual diagrams are finite , one can imagine situations where the two regularization methods give different partial results .
however , we do not observe significant cancellations in the sum of the differences .
thus , it seems the differences are real although practically negligible ; their sum is very small and amounts to @xmath65 , which is the error estimate in @xcite . thus our results agree within that error .
we have applied particle theory methods to compute , in dimensional regularization and a general covariant gauge , the corrections of order @xmath0 to the lamb shift .
we have made use of ibp techniques to reduce the problem of computing all the necessary feynman diagrams to the simpler problem of computing 32 scalar integrals .
mellin - barnes integral representations and sector decomposition have then allowed us to obtain analytic results for some of these integrals , and good numerical results for the rest . with this , we have been able to reproduce and improve the results from previous calculations .
the techniques used here are quite general and can be applied to other multi - loop problems in atomic physics .
we are grateful to a.v . and v.a .
smirnov for providing us with a new version of ` fiesta ` prior to publication .
we thank m.i .
eides for useful comments .
this work was supported by the natural sciences and engineering research council of canada .
the work of j.h.p . was supported by the alberta ingenuity foundation .
the feynman diagrams were drawn using ` axodraw ` @xcite and ` jaxodraw 2 ` @xcite .
in section [ eval ] we presented our method of calculating the corrections to the lamb shift , which differs significantly from the methods used in previous calculations .
one important difference is the reduction of diagrams to master integrals . here
we present our results for all master integrals . a graphic representation of the 32 master integrals .
solid and dashed lines represent massive and massless scalar propagators , respectively .
the dotted double line denotes the delta function .
a dot on a line signifies that the propagator is raised to a higher power .
the external lines indicate the momentum @xmath66 of the electron flowing in and out of the diagram .
the first two diagrams represent different integrals that differ only by a term in the numerator . ] the set of master integrals is represented in fig .
[ masters ] .
we have two different types of integrals .
in euclidean space , they are defined as : @xmath67^{\nu_{4}}}{\nonumber}\\ & & \times\frac{1}{[(k_{1}+k_{2}+p)^{2}+1]^{\nu_{5}}\,[(k_{1}+k_{2}+k_{3}+p)^{2}+1]^{\nu_{6}}\,[(k_{2}+k_{3}+p)^{2}+1]^{\nu_{7}}\,[(k_{3}+p)^{2}+1]^{\nu_{8}}}\,,\\ & & g(\nu_{1},\nu_{2},\nu_{3},\nu_{4},\nu_{5},\nu_{6},\nu_{7})=\frac{e^{3\gamma_{e}\epsilon}}{{\left(}\pi^{d/2}{\right)}^{3}}\int\frac{d^{d}k_{1}\ , d^{d}k_{2}\ , d^{d}k_{3}\;2\pi\,\delta(k_{1}^{0})}{(k_{1}^{2})^{\nu_{1}}\,(k_{2}^{2})^{\nu_{2}}\,(k_{3}^{2})^{\nu_{3}}\,[(k_{1}+k_{2}+p)^{2}+1]^{\nu_{4}}}{\nonumber}\\ & & \times\frac{1}{[(k_{1}+k_{2}+k_{3}+p)^{2}+1]^{\nu_{5}}\,[(k_{2}+k_{3}+p)^{2}+1]^{\nu_{6}}\,[(k_{3}+p)^{2}+1]^{\nu_{7}}}\,,\end{aligned}\ ] ] where @xmath68 , and @xmath69 is the momentum of the electron .
the mass of the electron has been set equal to one for convenience and can be easily restored from the dimension of the integral .
the factor @xmath70 , where @xmath71 is the euler - mascheroni constant , has been introduced to suppress the dependence of the results on this constant . with these definitions ,
our results for the master integrals are : @xmath72\,,\label{i11}\\ i_{12 } & = & f(1,1,0,0,1,1,0,1)=-\frac{4\pi^{2}}{\epsilon}-24\pi^{2}-\frac{4\pi^{4}}{3}+\pi^{2}{\left(}-116-\frac{59}{3}\pi^{2}+32\ln2 + 100\zeta(3){\right)}\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{13 } & = & f(1,1,0,1,0,1,0,1)=-263.74028719521945979(1)+1741.1125810306205720(1)\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{14 } & = & f(1,0,0,0,1,1,1,1)=-362.8560(1)+\mathcal{o}(\epsilon)\,,\\ i_{15 } & = & f(1,1,0,0,1,1,1,1)=36.969282(2)+\mathcal{o}(\epsilon)\,,\\ i_{16 } & = & f(1,4,0,0,1,0,1,1)=\pi^{2}\left(-\frac{513}{128\epsilon}-\frac{11077}{768}-\frac{571}{16}\ln2 + 32\sqrt{5}\ln\left(\frac{1+\sqrt{5}}{2}\right)\right){\nonumber}\\ & & -1889.3810189605726842(1)\epsilon-2199.2559561980712031(1)\epsilon^{2}+\mathcal{o}(\epsilon^{3})\,,\\ i_{17 } & = & f(1,0,0,0,2,0,2,1)=\frac{32\pi^{2}}{\sqrt{5}}\ln\left(\frac{1+\sqrt{5}}{2}\right)-683.43054120051764110(1)\epsilon{\nonumber}\\ & & + 5647.2496334930969112(1)\epsilon^{2}+\mathcal{o}(\epsilon^{3})\,,\\ i_{18 } & = & f(1,4,1,0,1,0,1,1)=\pi^{2}\left[\frac{343}{512\epsilon}+\frac{125257}{46080}-\frac{2}{3}\pi^{2}+\frac{169}{192}\ln2 + 16\ln^{2}2 - 48\ln^{2}{\left(}\frac{1+\sqrt{5}}{2}{\right)}\right]+\mathcal{o}(\epsilon)\,,\\ i_{19 } & = & f(1,0,1,0,1,1,1,0)=-293.4480(2)+\mathcal{o}(\epsilon)\,,\\ i_{20 } & = & f(1,0,0,1,1,0,1,1)=\pi^{2}\left[-\frac{2}{\epsilon}-2-\frac{8}{3}\pi^{2}+16\sqrt{5}\ln{\left(}\frac{1+\sqrt{5}}{2}{\right)}+32\ln^{2}{\left(}\frac{1+\sqrt{5}}{2}{\right)}\right]{\nonumber}\\ & & + 1394.0754186124348755(1)\epsilon+\mathcal{o}(\epsilon^{2})\,,\label{i20}\\ i_{21 } & = & f(1,1,1,0,0,1,0,1)=2\pi^{2}-\frac{4\pi^{4}}{3}+\pi^{2}{\left(}44 - 4\pi^{2}+80\zeta(3){\right)}\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{22 } & = & f(0,0,0,1,1,0,1,1)=-\frac{128\pi^{2}}{3}-\pi^{2}\left(\frac{1792}{3}+\frac{256}{3}\pi-\frac{1024}{3}\ln2\right)\epsilon+\mathcal{o}(\epsilon^{2})\,,\label{i22}\\ i_{23 } & = & f(1,0,1,0,1,0,1,0)\nonumber \\ & = & \frac{2\pi^{3}e^{3\gamma_{e}\epsilon}\,}{\gamma^{2}{\left(}2 - 2\epsilon{\right)}\gamma{\left(}\frac{3}{2}-\epsilon{\right)}}\left[2^{2\epsilon-2}\frac{\gamma{\left(}1-\epsilon{\right)}\gamma{\left(}\epsilon-\frac{1}{2}{\right)}}{\sin{\left(}\pi\epsilon{\right)}\cos{\left(}4\pi\epsilon{\right)}}{\left(}\frac{\gamma{\left(}\frac{7}{2}-5\epsilon{\right)}\sin{\left(}\pi\epsilon{\right)}}{\gamma{\left(}\frac{5}{2}-3\epsilon{\right)}\sin{\left(}2\pi\epsilon{\right)}\cos{\left(}3\pi\epsilon{\right)}}-\frac{\gamma{\left(}3\epsilon-\frac{3}{2}{\right)}}{\gamma{\left(}5\epsilon-\frac{5}{2}{\right)}\cos{\left(}2\pi\epsilon{\right)}}{\right)}\right.\nonumber \\ & & + \frac{\sqrt{\pi}\,\gamma{\left(}2 - 2\epsilon{\right)}\gamma{\left(}2\epsilon-\frac{1}{2}{\right)}}{\gamma{\left(}2-\epsilon{\right)}\gamma{\left(}3\epsilon-\frac{1}{2}{\right)}\sin{\left(}\pi\epsilon{\right)}\cos{\left(}\pi\epsilon{\right)}\cos{\left(}3\pi\epsilon{\right)}}\,{}_{3}f_{2}{\left(}1,2 - 2\epsilon,2\epsilon-\frac{1}{2};2-\epsilon,3\epsilon-\frac{1}{2};1{\right)}\nonumber \\ & & \left.-\frac{2^{1 - 2\epsilon}\pi\,\gamma{\left(}\frac{5}{2}-3\epsilon{\right)}}{\gamma{\left(}\frac{5}{2}-2\epsilon{\right)}\gamma{\left(}\frac{1}{2}+\epsilon{\right)}\sin{\left(}2\pi\epsilon{\right)}\cos{\left(}\pi\epsilon{\right)}\cos{\left(}2\pi\epsilon{\right)}}\,{}_{3}f_{2}{\left(}1,\frac{5}{2}-3\epsilon,\epsilon;\frac{5}{2}-2\epsilon,2\epsilon;1{\right)}\right]\,,\\ i_{24 } & = & g(0,1,2,1,0,1,0)=\frac{2\pi^{2}}{\epsilon}-162.745878930257(1)+640.681562239(2)\epsilon{\nonumber}\\ & & -9490.745115169417(3)\epsilon^{2}+\mathcal{o}(\epsilon^{3})\,,\\ i_{25 } & = & g(2,1,1,1,0,1,0)=-\frac{4\pi^{2}}{\epsilon}-192.3546921335253(1)-2297.18352848038(1)\epsilon{\nonumber}\\ & & -10356.58582995624(1)\epsilon^{2}+\mathcal{o}(\epsilon^{3})\,,\\ i_{26 } & = & g(1,1,1,1,0,1,0)=-\frac{4\pi^{2}}{\epsilon}-244.4995291143211(3)-2339.54007847666(2)\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{27 } & = & g(0,1,1,2,0,1,1)=136.8086023(2)-907.048(2)\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{28 } & = & g(0,1,1,1,1,1,0)=-280.62418(1)+734.494(1)\epsilon+\mathcal{o}(\epsilon^{2})\,,\\ i_{29 } & = & g(1,1,1,0,1,1,1)=118.63826101784(1)+\mathcal{o}(\epsilon)\,,\\ i_{30 } & = & g(0,1,1,0,1,1,0)=-10\sqrt{\pi}e^{3\gamma_{e}\epsilon}\,\frac{\gamma\left(\epsilon\right)\gamma^{2}{\left(}1-\epsilon{\right)}\gamma\left(\frac{5}{2}-5\epsilon\right)\gamma{\left(}-\frac{3}{2}+3\epsilon{\right)}}{\gamma{\left(}2 - 2\epsilon{\right)}\gamma\left(\frac{7}{2}-4\epsilon\right)}{\nonumber}\\ & & \times{}_{3}f_{2}\left(\frac{7}{2}-5\epsilon,\frac{3}{2}-\epsilon,-\frac{1}{2}+\epsilon;\frac{7}{2}-4\epsilon,\frac{1}{2}+\epsilon;1\right)\,,\\ i_{31 } & = & g(1,1,1,1,1,1,0)=49.3616(1)+\mathcal{o}(\epsilon)\,,\\ i_{32 } & = & g(1,1,1,1,1,1,1)=26.272804(6)+291.1097(1)\epsilon+\mathcal{o}(\epsilon^{2})\,,\end{aligned}\ ] ] where @xmath73 denotes riemann s zeta function , and @xmath74 is a generalized hypergeometric function .
the latter can be expanded in @xmath75 with the help of the ` mathematica ` package ` hypexp 2 ` @xcite .
the relation between integrals @xmath76 and @xmath77 expressed in eq .
( [ i2 ] ) is not evident when looking at their respective representations .
this relation becomes clear when checking the cancellation of the gauge - parameter dependence in the sum of all diagrams .
if the 32 integrals presented here were an irreducible basis , the gauge dependence of the coefficient of each integral should vanish independently .
however , this does not happen with the coefficients of integrals @xmath76 and @xmath77 , which means that the integrals are connected . demanding the cancellation of the gauge dependence yields eq .
( [ i2 ] ) .
we checked this relation by computing explicitly the analytic solution for @xmath77 .
there appears to be also a relation between integrals @xmath78 and @xmath79 , although the gauge dependence does not give us any hint in this case . by demanding the cancellation of poles in several diagrams
, one can find the following relation between the first three terms of @xmath78 and @xmath79 , @xmath80 the relation , however , seems to be valid to all orders in the @xmath75 expansion .
we checked it numerically up to order @xmath81 , but we could not find an analytic proof for it .
integrals @xmath78@xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath86 , and @xmath87@xmath88 can be represented as a one - fold mellin - barnes integral .
we only show numerical results with 20-digit precision , which is more than enough for our purposes .
however , these integrals can be easily evaluated with a precision of 100 digits or more . with this kind of precision it is possible to find analytical results , using the ` pslq ` algorithm @xcite . in this way , we determined the @xmath89 term of integrals @xmath83 and @xmath90 .
the analytic expression for the @xmath91 term in @xmath87 was obtained using the analytic result for set ii of vacuum - polarization diagrams presented in @xcite .
likewise , the analytic expression for the @xmath89 term in @xmath88 was extracted from the analytic result for set iii found in @xcite . as mentioned above
, we were able to numerically calculate these integrals to 100-digit precision and confirm the analytic expressions with ` pslq ` .
here we show the analytic results for diagrams @xmath56 and @xmath57 from fig .
[ other ] : @xmath92 k. pachucki , phys . rev . a * 63 * , 042503 ( 2001 ) ; m. i. eides and v. a. shelyuto , can . j. phys . * 85 * , 509 ( 2007 ) [ arxiv : physics/0612244 ] .
k. pachucki , phys .
a * 53 * , 2092 ( 1996 ) .
b. lauss , nucl .
a * 827 * , 401c ( 2009 ) [ arxiv:0902.3231 [ nucl - ex ] ] ; a. antognini _ et al_. , aip conf
. proc . * 796 * , 253 ( 2005 ) .
k. pachucki , phys .
lett . * 72 * , 3154 ( 1994 ) m. i. eides and v. a. shelyuto , pisma zh .
eksp . teor
. fiz .
* 61 * , 465 ( 1995 ) [ jetp lett .
* 61 * , 478 ( 1995 ) ] . m. i. eides and v. a. shelyuto , phys .
a * 52 * , 954 ( 1995 ) [ arxiv : hep - ph/9501303 ] . for a quantum field
theoretical treatment of non - relativistic bound states see , for example , sec .
5.3 of m. e. peskin and d. v. schroeder , _ quantum field theory _ , ( westview press , boulder , 1995 ) .
p. nogueira , j. comput .
phys . * 105 * , 279 ( 1993 ) .
r. harlander , t. seidensticker and m. steinhauser , phys .
b * 426 * , 125 ( 1998 ) [ arxiv : hep - ph/9712228 ] .
t. seidensticker , arxiv : hep - ph/9905298 . j. a. m. vermaseren , arxiv : math - ph/0010025 .
m. steinhauser , comput .
commun . * 134 * , 335 ( 2001 ) [ arxiv : hep - ph/0009029 ] ; url : http://www-ttp.particle.uni-karlsruhe.de/~ms/software.html .
f. v. tkachov , phys .
b * 100 * , 65 ( 1981 ) ; k. g. chetyrkin and f. v. tkachov , nucl .
b * 192 * , 159 ( 1981 ) .
s. laporta and e. remiddi , phys .
b * 379 * , 283 ( 1996 ) [ arxiv : hep - ph/9602417 ] .
s. laporta , int .
j. mod .
a * 15 * , 5087 ( 2000 ) [ arxiv : hep - ph/0102033 ] .
a. v. smirnov , jhep * 0810 * , 107 ( 2008 ) [ arxiv:0807.3243 [ hep - ph ] ] .
v. a. smirnov , phys .
b * 460 * , 397 ( 1999 ) [ arxiv : hep - ph/9905323 ] .
j. b. tausk , phys .
b * 469 * , 225 ( 1999 ) [ arxiv : hep - ph/9909506 ] .
m. czakon , comput .
commun . * 175 * , 559 ( 2006 ) [ arxiv : hep - ph/0511200 ] .
a. v. smirnov and v. a. smirnov , eur .
j. c * 62 * , 445 ( 2009 ) [ arxiv:0901.0386 [ hep - ph ] ] .
a. v. smirnov and m. n. tentyukov , comput .
commun . *
180 * , 735 ( 2009 ) [ arxiv:0807.4129 [ hep - ph ] ] .
t. hahn , comput .
. commun .
* 168 * , 78 ( 2005 ) [ arxiv : hep - ph/0404043 ] .
t. binoth and g. heinrich , nucl .
b * 585 * , 741 ( 2000 ) [ arxiv : hep - ph/0004013 ] ; nucl .
b * 680 * , 375 ( 2004 ) [ arxiv : hep - ph/0305234 ] ; nucl .
b * 693 * , 134 ( 2004 ) [ arxiv : hep - ph/0402265 ] .
g. heinrich , int .
j. mod .
a * 23 * , 1457 ( 2008 ) [ arxiv:0803.4177 [ hep - ph ] ] .
a. v. smirnov , v. a. smirnov and m. tentyukov , arxiv:0912.0158 [ hep - ph ] .
k. pachucki , phys .
a * 48 * , 2609 ( 1993 ) . h. m. fried and d. r. yennie , phys .
112 * , 1391 ( 1958 ) . g. s. adkins , phys .
d * 47 * , 3647 ( 1993 ) .
m. i. eides , h. grotch and v. a. shelyuto , phys .
a * 55 * , 2447 ( 1997 ) [ arxiv : hep - ph/9610443 ] . j. a. m. vermaseren , comput . phys
. commun .
* 83 * , 45 ( 1994 ) .
d. binosi , j. collins , c. kaufhold and l. theussl , comput .
commun . * 180 * , 1709 ( 2009 ) [ arxiv:0811.4113 [ hep - ph ] ] .
t. huber and d. matre , comput . phys
. commun . * 178 * , 755 ( 2008 ) [ arxiv:0708.2443 [ hep - ph ] ] .
ferguson , d.h .
bailey and s. arno , math .
comput . * 68 * , 351 ( 1999 ) . | we present results for the corrections of order @xmath0 to the lamb shift .
we compute all the contributing feynman diagrams in dimensional regularization and a general covariant gauge using a mixture of analytical and numerical methods .
we confirm results obtained by other groups and improve their precision .
values of the 32 `` master integrals '' for this and similar problems are provided . |
in the attempt to extract values of physical parameters from stellar spectra one first tries to determine the value of radial velocity .
if the spectra have a wide wavelength coverage , like in the case of elodie spectra , it can be assumed that a fixed spectral mask gives radial velocities with sufficient accuracy ( katz et al .
analysis of gaia radial velocity spectrometer ( rvs ) data will profit from quite accurate knowledge of stellar parameters obtained from on - board photometric observations .
so all simulations of gaia radial velocity determination assumed that the choice of a proper template is given by complimentary photometric data ( katz et al .
2004 , munari et al .
2003 , zwitter 2002 ) .
alternatively it was assumed , while assessing the capabilities of gaia spectra to independently determine the values of stellar parameters , that the radial velocity is a - priori known ( esa-2000-sci-4 , thevenin et al .
2003 , katz et al .
note that these approaches are relevant at the end of the gaia mission , but at the start of the mission virtually nothing will be known about the values of stellar parameters , so that one would not know which kind of template to use for a radial velocity determination .
also , the wavelength coverage of gaia spectra is much smaller than that of a typical echelle ( elodie ) type spectrum .
so we attempt here to determine _ both _ the best radial velocity and the values of parameters of the stellar atmosphere simultaneously .
data from the rave experiment , which are very similar in both wavelength coverage and resolution to those of gaia s rvs , will be used to illustrate the point .
note that since virtually nothing is known about the rave targets it is obvious to use such an approach .
rave ( radial velocity experiment ) project ( steinmetz 2003 , munari et al .
2004 ) is an ambitious international collaboration aimed at conducting an all - sky spectroscopic survey of galactic stars .
the participating countries are australia , canada , france , germany , italy , japan , netherlands , slovenia , switzerland , uk , and usa . the spectra allow us to measure radial velocities as well as metallicites and other stellar parameters for a large number of stars using the 1.2-m uk schmidt telescope at the anglo - australian observatory .
data collection for this project started in april 2003 with more than 40,000 spectra observed to date .
spectral resolving power and wavelength range ( 84138746 ) are very similar to those of the gaia radial velocity spectrograph .
observed or synthetic spectra can be used for classification .
the advantage of the former is that they are `` real '' , so avoiding any simplifying assumptions used in synthetic spectra calculation ( munari et al .
the latter have the advantage of a more uniform coverage of the parameter space . at present
we are using a subset of a synthetic library of nearly 60.000 spectra with suitable resolving power ( zwitter , castelli & munari 2004 ) which were calculated from the latest generation of kurucz models .
the grid is characterized by the following ranges of stellar parameters : @xmath0 , @xmath1 , @xmath2 \le + 0.5 $ ] , @xmath3 km s@xmath4 , @xmath5 km s@xmath4 .
correlation routines , like the iraf s rvsao package ( kurtz & mink 1998 ) are a well established way to calculate radial velocity using a fixed stellar or galaxy spectrum template . their figure of merit is based on the value of correlation coefficient @xmath6 ( tonry & davis 1979 ) .
but one should not conclude that the template spectrum with the highest value of @xmath6 also gives the best fit to the observed spectrum .
1 shows results of analysis of a single observed spectrum using a set of template spectra .
the maximal value of the correlation coefficient @xmath6 occurs at the `` stem '' of area covering a caravel ship shape , though the minimum of @xmath7 occurs at its bottom .
the reason is that the correlation is defined using normalized quantities .
therefore it is not sensitive to any kind of constant shift or spread ( of the kind @xmath8 and @xmath9 where @xmath10 is a constant value or bias ) .
this explains why it is only sensitive to high variations in the line profile and is not suitable for minimization .
we can also say that in cross - correlation one tries to find the best alignment of the lines , in a way irrespective of their depth .
this is strictly valid for isolated single lines .
the effective wavelength of a blend of nearby lines can shift sideways and so worsens the cross - correlation only with a large change of the temperature and gravity .
figure 2 plots the observed spectrum which was used in figure 1 and one of the templates from the library .
the match is poor even though the template shown features the highest value of the correlation coefficient @xmath6 among all templates in the library ( the point is at the tip of the stem of the caravel shape in figure 1 ) .
cross - correlation alone is clearly not enough to find the template with the best match .
the situation can be improved by supplementing correlation analysis with a proper minimization scheme ( figure 3 ) .
the reason is that normalized squared distance is sensitive to line depth and shape as well as continuum level , and so to the values of stellar parameters . in our case
we first use a correlation routine to calculate the velocity shift between the observed spectrum and a template from the synthetic library .
but the goodness of fit is not based on the value of the correlation coefficient r. instead we calculate the standard deviation between the observed spectrum and the velocity - shifted template . the template which yields the smallest standard deviation
is assumed to give a fair representation of parameters of the stellar atmosphere .
values of several stellar parameters are primarily based on spectral analysis .
these include metallicity , abundances of individual elements and stellar rotation velocity .
the values of effective temperature and surface gravity can also be derived , or in the case of gaia checked , by spectral analysis .
both gaia and rave projects will provide too many spectra to be analyzed by hand .
automated procedures , such as the one proposed here , yield useful results .
any other information available , in the case of gaia these would be results of photometric and astrometric observations , can be seen as an additional constraint on the relevant range of stellar templates to be considered for a given star .
if an acceptable match could not be found despite a high s / n of the observed spectrum , this implies that any kind of peculiarity is present in the observed object , or that the observed star is a double lined spectroscopic binary .
cross correlation is sensitive to position of the lines and so radial velocity .
it is however not sensitive to line depth , so it is not to be used to determine the values of stellar atmosphere parameters unless blends of nearby lines are driving the correlation .
the situation gets worse with low s / n spectra and when sharp lines ( e.g. ca ii ir triplet ) are present . by adding a suitable minimum distance scheme which is sensitive to equivalent widths ( thevenin & foy 1983 ) the values of both radial velocity and stellar parameters like temperature , gravity , metallicity and rotation velocity can be determined reliably
tz acknowledges financial support from the slovenian ministry for education , science and sports .
katz , d. , soubiran , c. , cayrel , r. , adda , m. , cautain , r. 1998 , a&a 338 , 151 katz , d. , et al .
2004 , mnras , in press kurtz , m.j . , mink , d.j .
1998 , pasp 110 , 934 munari , u. , agnolin , p. , tomasella , l. 2001 , balt.a .
10 , 613 munari , u. , zwitter , t. , katz , d. , cropper , m. 2003 , in gaia spectroscopy : science and technology , asp conf .
298 , 275 munari , u. , zwitter , t. , siebert , a. 2004 , this volume steinmetz , m. , 2003 , in gaia spectroscopy : science and technology , asp conf .
298 , 381 thevenin , f. , foy , r. 1983 , a&a 122 , 261 thevenin , f. , bijaoui , a. , katz , d. 2003 , in gaia spectroscopy : science and technology , asp conf .
298 , 291 tonry , j. , davis , m. 1979 , aj 84 , 1511 zwitter , t. 2002 , a&a 386 , 748 zwitter , t. , castelli , f. , munari , u. 2004 , a&a , 417 , 1055 | one may consider cross - correlation routines applied to stellar spectra as a way to determine not only the radial velocities but also to assess the properties of stellar atmospheres .
we show that this is not the case . high value of the cross - correlation coefficient between the observed and a template spectrum does not mean that the template corresponds to correct values of stellar temperature , gravity or metallicity .
cross - correlation operates in normalized space , so it is not sensitive to scaling or shifting of stellar flux .
it is argued that cross - correlation is useful to efficiently determine the velocity shift between the observed and the template spectrum . but minimal distance methods need to be used to determine which template gives the best match to the observed spectrum , and so derive values of physical parameters . to illustrate the point we use data from the rave project which are being obtained at very similar spectral resolution and wavelength domain as will be the case for gaia s rvs .
[ 2004/06/22 1.2 ( kof ) ; 2001/04/25 1.1 ( pwd ) ] |
18483 was discovered in 2003 during deep observations of the galactic center @xcite .
the mean source x - ray flux was @xmath410 mcrab in the 15 - 40 kev , and @xmath45 mcrab in the 40 - 100 kev band @xcite .
the 18.5 days orbital period of the system was discovered by @xcite using _ rxte _ archival data , and was later confirmed with @xcite .
data also showed that 18483 usually undergoes relatively long outbursts ( @xmath43 days ) that comprise several fast flares with typical timescales of a few hours . during these bright events , the broad band
( 3 - 50 kev ) spectrum is best fit by an absorbed cutoff power law model ( photon index @xmath5=1.4 , cut - off energy @xmath6=22 kev , and absorption column density @xmath7=9@xmath110@xmath8 @xmath9 ) . @xcite
further detected a periodicity at [email protected] s in the data , and interpreted it as the spin - period of the neutron star ( ns ) hosted in 18483 .
the measured pulse fraction in the 4 - 20 kev energy band was 48@xmath07%=(@xmath10-@xmath11)/(@xmath10+@xmath11 ) , where @xmath10 and @xmath11 are the measured count rates at the maximum and at the minimum of the folded light curve , respectively . ] .
/xrt observations in 2006 detected the source at a flux level of 4.2@xmath110@xmath12 erg/@xmath13/s and provided a source position of @xmath14=18@xmath1548@xmath1617@xmath1717 and @xmath18=-3@xmath1910@xmath2015@xmath2154 ( estimated accuracy 3@xmath213 , * ? ? ?
this allowed @xcite to identify the optical counterpart of the x - ray source , a b0.5ia star at a distance of 3 - 4 kpc , and to estimate its mass and radius ( @xmath22=33 @xmath23 and @xmath24=33.8 @xmath25 , respectively ) .
these authors also suggested that an eccentricity 0.43@xmath26@[email protected] could explain the 3-day duration of the outbursts ( as reported by * ? ? ? * ) .
based on these results , it was concluded that 18483 most likely belongs to the class of supergiant fast x - ray transients ( sfxts , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , due to the longer duration of its outbursts ( a few days as opposed to a few hours ) and a factor of @xmath410 lower luminosity swings between outburst and quiescence , @xcite classified 18483 as an `` intermediate '' sfxt , rather than a standard sfxt ( see * ? ? ?
* for the definition of standard and intermediate sfxts ) . in this paper
we analyze a 18 ks observation of 18483 in quiescence , and report the results of the spectral and timing analysis of this observation .
we found that the pulse fraction of the source x - ray emission decreased significantly with respect to that measured while the source was in outburst , and provide for the first time an estimate of the spin - period derivative of this source .
we also analyzed all the available /xrt observations of 18483 , and studied the orbital variations of the source x - ray flux .
a 1 ks observation is also analyzed and provided an improved position of the x - ray source .
the results from this study are then compared with those obtained recently on 18 , the other sfxt with a similar orbital period to that of 18483 .
so far , the orbital period has been measured with certainty only in other two sfxts , i.e. igrj16479 - 4514 ( 3.3194 d * ? ? ?
* ) , and igrj11215 - 5952 @xcite .
( @xcite ) observed 18483 on 2006 october 12 , and the total good exposure time was 14.4 ks ( we discarded observational intervals that were affected by a high background ) .
the observation data files ( odfs ) were processed to produce calibrated event lists using the standard science analysis system ( sas 8.0 ) .
we used the epproc and emproc tasks for the epic - pn and the two mos cameras , respectively .
source light curves and spectra were extracted in the 0.510 kev band , by using a circular extraction region with a radius of 20 .
background light curves and spectra were instead extracted by using a circular region with a radius of 50 .
we used the sas backscale task and the lcmath task in heasoft ( version 6.6.1 ) to account for the difference in extraction areas between source and background spectra and light curves , respectively .
the times of all light curves were corrected to the barycentre of the solar system with the sas barycen task . in all cases , owing to poor statistics
, the epic - mos1 and epic - mos2 cameras did not contribute significant additional information on the source spectra .
therefore , in the following we discuss only the spectra from the epic - pn camera .
all spectra were rebinned in order to have at least 25 photons for each energy bin . in fig .
[ fig : lcurve ] we report the x - ray light curves of the source in the 0.5 - 5 kev and 5 - 10 kev energy bands ; the lower panel of the figure shows the hardness ratio ( i.e. the ratio of the count - rate in the hard , 5 - 10 kev , and soft , 0.5 - 5 kev , energy band ) versus time .
we note that the source count rate was decreasing during the first 5 ks of the observation .
unfortunately , the number of counts was insufficient to carry out any detailed investigation of the spectral variability .
therefore , we extracted only the 0.5 - 10 kev spectrum by using the total exposure time of the observation , and performed a fit with an absorbed power law model .
the best fit parameters were @xmath7=7.7@xmath28@xmath110@xmath8 @xmath9 , and @[email protected] ( @xmath29/d.o.f.=1.3/39 ; hereafter errors are at 90% confidence level , unless otherwise indicated ) .
the absorbed ( unabsorbed ) flux in the 0.5 - 10 kev band was 9.3@xmath110@xmath30 erg/@xmath13/s ( 5.2@xmath110@xmath12 erg/@xmath13/s ) . assuming a source distance of 4 kpc , the unabsorbed flux corresponds to an x - ray luminosity ( 0.5 - 10 kev ) of 1.0@xmath110@xmath31 erg / s .
figure [ fig : spectrum ] shows the epic - pn source spectrum , together with the best fit model and the residuals from this fit .
the 90% confidence upper limit to the equivalent width for narrow iron lines at 6.4 kev and 6.7 kev is 0.13 kev and 0.10 kev , respectively .
timing analysis of the data was carried out by using barycentred event files .
we searched for the 21.0526 s spin - period of the ns in 18483 by performing first a power spectrum of the data .
no significant evidence for a peak at the corresponding frequency was found . in order to investigate further the presence of pulsations in the data
, we applied the z@xmath32-statistic technique @xcite to the photons event distribution for trial frequencies in a small window centered on 0.0475 hz @xcite .
a spin - period of [email protected] s ( hereafter errors on the ns spin periods are all at 1@xmath33 confidence level ) is found with a peak power of z@xmath32@xmath426 .
the single - trial significance of this period is 4.7 @xmath33 .
figure [ fig : zstat ] shows the power spectrum computed with the z@xmath32-statistic technique by using the total exposure time of the observation .
this period estimate was then refined by employing a phase fitting technique ( see e.g. , * ? ? ?
this gave our best determined spin period of @[email protected] s. in order to derive the significance of this result over the entire range of spin periods considered , we assumed a spin period derivative of 1.3@xmath110@xmath2 s / s ( see sect .
[ sec : discussion ] ) and multiplied the single trial significance of the z@xmath32-statistic for the total number of trial @xmath35/(@xmath36/2@xmath37 ) . here
@xmath35 is the separation in seconds between our measured spin period and that reported by @xcite , and @xmath37=14.4 ks is the total duration of the observation .
this gave us a significance of 3.7@xmath33 .
from the folded light curve of the observation ( obtained with the efold task , see fig .
[ fig : phase ] ) , we measured a pulsed fraction of f=15@xmath03% in the 0.5 - 10 kev energy band .
the profile is consistent with a sinusoid . in order to investigate the energy dependence of the pulse fraction
, we also extracted and folded light curves in different energy bands and time intervals .
we found that the pulsed fraction decreases slightly toward higher energies , whereas no significant variation could be measured across the `` a '' and `` b '' time intervals shown in fig [ fig : lcurve ] ( see table [ tab : pulse ] ) .
we note that the pulsed fraction we measured in the quiescent state of 18483 is a factor of @xmath43 lower than that reported by @xcite during the source outburst ) . ]
. @llll@ energy band & total obs . &
a & b + 0.5 - 10 kev & ( 15@xmath03)@xmath38 & @xmath3929@xmath38 @xmath40 & @xmath3932@xmath38 @xmath40 + 0.5 - 5 kev & ( 23@xmath03)@xmath38 & ( 30@xmath05)@xmath38 & @xmath3933@xmath38 @xmath40 + 5 - 10 kev & @xmath3918@xmath38 @xmath40 & @xmath3936@xmath38 @xmath40 & @xmath3929@xmath38 @xmath40 + + [ tab : pulse ] @lllllllll@ igrj18483 - 0311 + obs i d & instr & start time & stop time & exp & @xmath7 & @xmath5 & @xmath41 & @xmath29/d.o.f . + & & & & ks & ( 10@xmath8 @xmath9 ) & & ( erg/@xmath42/s ) + 00035093001@xmath43 & xrt / pc & 2006 - 02 - 16 01:37:19 & 2006 - 02 - 16 22:36:57 & 7.9 & 6.0@xmath44 & 1.4@xmath45 & 5.3@xmath110@xmath46 & 1.3/21 + 00035093002@xmath43 & xrt / pc & 2006 - 03 - 05 11:12:39 & 2006 - 03 - 05 17:58:56 & 5.6 & [email protected] & [email protected] & 6.4@xmath110@xmath46 & 1.1/51 + 00035093003 & xrt / pc & 2008 - 09 - 26 13:49:39 & 2008 - 09 - 26 15:26:38 & 2.0 & 5.0 ( fixed ) & 1.2 ( fixed ) & 1.8@xmath110@xmath12 & + saxj1818.6 - 1703 + 00036128001 & xrt / pc & 2007 - 11 - 09 17:47:09 & 2007 - 11 - 09 21:04:57 & 1.6 & 6.0 ( fixed ) & 1.9 ( fixed ) & @xmath392.1@xmath110@xmath12 @xmath47 & + 00036128003 & xrt / pc & 2008 - 04 - 18 14:38:56 & 2008 - 04 - 18 17:45:49 & 2.0 & 6.0 ( fixed ) & 1.9 ( fixed ) & 3.5@xmath110@xmath12 @xmath48 & + 00037889001 & xrt / pc & 2008 - 07 - 20 02:44:06 & 2008 - 07 - 21 01:21:56 & 3.7 & 6.0 ( fixed ) & 1.9 ( fixed ) & 8.5@xmath110@xmath12 @xmath48 & + + + + + [ tab : log ] in table [ tab : log ] we show a log of the observations analyzed in the present study .
note that the observations i d 00035093001 and 00035093002 were also published previously by @xcite .
we used the xrtpipeline ( v.0.12.1 ) task to process /xrt data ( note that part of these data were published by * ? ? ?
standard event grades of 0 - 12 were selected for the xrt photon - counting ( pc ) mode ; filtering and selection criteria were applied using ftools ( heasoft v.6.6.1 ) .
we created exposure maps through the xrtexpomap task , and used the latest spectral redistribution matrices in the heasarc calibration database ( v.011 ) .
ancillary response files , accounting for different extraction regions , vignetting and psf corrections , were generated using the xrtmkarf task . when required , we corrected pc data for pile - up , and used the xrtlccorr to account for this correction in the background subtracted light curves . for each observation in table [
tab : log ] , we extracted the light curve and spectrum , and derived a mean x - ray flux by fitting this spectrum with an absorbed power - law model ( we used xspec v.12.5.0 ) .
spectra were rebinned in order to have at least 20 photons per bin and allow for @xmath49 fitting . in the observation
00035093003 the very low source count rate ( [email protected]@xmath110@xmath50 ) did not allow for a detailed spectral analysis .
therefore , we estimated the source count rate of the observation with sosta ( ximage v.4.4.1 ) , and then used this count rate within webpimms in order to derive the x - ray flux ( we assumed the same spectral model of the observation i d .
00035093002 ) .
in table [ tab : log ] , we report the best fit parameters obtained using an absorbed power law model to characterize the source spectra ; fig.[fig : lcurve_swift ] shows the source light curves from the observations i d . 00035093001 and 00035093002 ( note that , owing to poor statistics , we do not report the light curve of the observation i d .
00035093003 ) .
@lllll@ igrj18483 - 0311 & & & + instrument & phase & @xmath41@xmath40 & @xmath51@xmath52 & energy + & & ( erg/@xmath42/s ) & ( erg / s ) & range + ibis / isgri@xmath53 & 0 & 1.1@xmath110@xmath2 & 2.1@xmath110@xmath54 & 20 - 100 + swift / xrt@xmath55 & 0.13 & 1.8@xmath110@xmath12 & 3.5@xmath110@xmath56 & 0.5 - 10 + chandra / hrc@xmath55 & 0.24 & 4.3@xmath110@xmath57 & 8.3@xmath110@xmath58 & 0.5 - 10 + xmm / epic - pn@xmath55 & 0.52 & 5.2@xmath110@xmath12 & 1.0@xmath110@xmath31 & 0.5 - 10 + swift / xrt@xmath55 & 0.59 & 6.4@xmath110@xmath46 & 1.2@xmath110@xmath58 & 0.5 - 10 + swift / xrt@xmath55 & 0.67 & 5.3@xmath110@xmath46 & 1.0@xmath110@xmath58 & 0.5 - 10 + saxj1818.6 - 1703 & & & + instrument & phase & @xmath41@xmath40 & @xmath51@xmath59 & energy + & & ( erg/@xmath42/s ) & ( erg / s ) & range + ibis / isgri@xmath60 & 0 & 3.8@xmath110@xmath57 & 2.9@xmath110@xmath58 & 18 - 60 + ibis / isgri@xmath61 & 0.01 & 3.0@xmath110@xmath2 & 2.3@xmath110@xmath54 & 18 - 60 + ibis / isgri@xmath62 & 0.06 & 3.8@xmath110@xmath57 & 2.9@xmath110@xmath58 & 18 - 60 + sax / wfc@xmath63 & 0.07 & 2.1@xmath110@xmath2 & 1.6@xmath110@xmath54 & 2 - 9 + swift / xrt@xmath55 & 0.12 & 3.5@xmath110@xmath12 & 2.6@xmath110@xmath56 & 0.5 - 10 + swift / xrt@xmath55 & 0.22 & 8.5@xmath110@xmath12 & 6.4@xmath110@xmath56 & 0.5 - 10 + xmm / epic - pn@xmath64 & 0.51 & @xmath391.1@xmath110@xmath30@xmath65 & @xmath398.3@xmath110@xmath66@xmath65 & 0.5 - 10 + swift / xrt@xmath55 & 0.76 & @xmath392.1@xmath110@xmath12@xmath65 & @xmath391.6@xmath110@xmath56@xmath65 & 0.5 - 10 + chandra / acis - s@xmath67 & 0.89 & 7.5@xmath110@xmath12 & 5.7@xmath110@xmath56 & 0.5 - 10 + ibis / isgri@xmath68 & 0.91 & 3.0@xmath110@xmath57 & 2.3@xmath110@xmath58 & 18 - 45 + ibis / isgri+swift / bat@xmath60 & 0.98 & @xmath393.8@xmath110@xmath57@xmath65 & @xmath392.9@xmath110@xmath58@xmath65 & 18 - 60 + ibis / isgri@xmath60 & 1.00 & @xmath393.8@xmath110@xmath57@xmath65 & @xmath392.9@xmath110@xmath58@xmath65 & 18 - 60 + ibis / isgri+swift / bat@xmath60 & 1.00 & 9.1@xmath110@xmath57 & 6.9@xmath110@xmath58 & 18 - 60 + swift / bat@xmath69 & 1.00 & 2.2@xmath110@xmath2 & 1.7@xmath110@xmath54 & 15 - 150 + + + + + + + + + + + + + + + [ tab : comparison ] observed 18483 on 2008 february 19 for a total exposure time of 1.2 ks with the high resolution camera .
we reduced these data using the ciao software ( v 4.1.1 ) and the latest calibration file available .
the best source position is provided by the wavdetect task at @xmath14=18@xmath1548@xmath1617@xmath172 and @xmath18=-3@xmath1910@xmath2016@xmath218 ( the position accuracy is 0@xmath218 at a 90% confidence level ) .
this is perfectly in agreement with the optical position reported by @xcite .
we also derived the source count rate ( [email protected] cts / s ) of the observation and then used this count rate with webpimms in order to estimate the source x - ray flux .
the results are given in table [ tab : comparison ] ( we assumed the same spectral model of the observation ) .
18 is the only sfxt whit an orbital period comparable to that of 18483 ( [email protected] days , * ? ? ?
* ; * ? ? ?
* ) . in table
[ tab : comparison ] we report all observations of this source we found in the literature , together with two recent /xrt observations that have not yet been published ( i d . 00036128001 and 00037889001 ) .
we analyzed all these /xrt observations with the procedures described in sect .
[ sec : swift ] and reported the results in table [ tab : log ] .
following @xcite , we measured in table [ tab : comparison ] the orbital phase of each observation from the epoch of the outburst occurred on 53671 mjd ( phase 0 ) , so as to permit a comparison with the orbital changes of the x - ray flux in 18483 .
this comparison is carried out in sect .
[ sec : discussion ] .
in this paper we reported on all available quiescent observations of 18483 , one of the two sfxts for which the spin and orbital periods have been measured with certainty ( the other is igrj11215 - 5952 with @xmath34=186.78 s , but see sect . [ sec : intro ] ) .
we report the detection of pulsations in the quiescent x - ray emission of this source , and give for the first time a measurement of its spin - period derivative . to our knowledge ,
the spin - period has so far been detected unambiguously in two other sfxts ( @xmath34=228 s , and 4.7 s in igrj16465 - 4507 and igrj1841.0 - 0536 , respectively ; * ? ? ?
* ; * ? ? ?
* ) ; however , the orbital period of these sources is not known . on the contrary , in the case of 18 the orbital period
is known , but the spin - period remains to be discovered .
recently it has been suggested that a measurement of the ns spin and orbital periods can be the key to distinguish between different models proposed for sfxt sources @xcite .
in fact , all these models involve a ns accreting from the intense wind of its supergiant companion , but several different mechanisms have been invoked in order to explain the very large luminosity swings observed during their transitions between outburst and quiescence @xcite .
in particular , @xcite suggested that , if very slow spinning nss ( @xmath34@xmath701000 s ) in relatively close orbits ( few tens of days ) are hosted in sfxts , then a magnetic gating mechanism can be invoked in order to explain such luminosity swings . in this case
the ns magnetic field would be in the `` magnetar '' range ( i.e. @xmath7010@xmath71 - 10@xmath72 g ; * ? ? ?
on the contrary , faster spin - periods might indicate that the large luminosity swings of sfxts are caused by a centrifugal rather than a magnetic gating ( a similar mechanism was suggested to explain the pronounced activity of be x - ray pulsar transient systems ; * ? ? ? * ) .
alternatively , the observed variations in the x - ray luminosity of sfxts might also be caused by drastic changes in the mass accretion rate onto the ns due to an extremely clumpy wind or to large scale structure in the immediate surroundings of the supergiant companion . in these models
, the orbital periods may be as high as hundreds of days ( see in particular * ? ? ? * ; * ? ? ?
* ) . in 2008 ,
an observation of igrj16479 - 4514 revealed that also eclipse - like events can contribute to the luminosity swings observed in sfxts @xcite .
therefore , besides a measurement of the ns spin and orbital period , also an in - depth monitoring of the x - ray flux and spectral changes at different orbital phases is required in order to distinguish between different models or scenarios for sfxt sources . to this aim , we presented in table [ tab : comparison ] an analysis of the orbital changes in the x - ray flux observed from 18483 and 18 , the only two sfxts with a comparable orbital period .
in the case of 18483 only few observations have been carried out in quiescence and thus the orbital monitoring of this source is far from being complete ( following * ? ? ?
* we measured the source phases from the epoch of the brightest outburst observed with at 53844.2 mjd ) .
the lowest flux state of this source was caught by /xrt at phase 0.13 , i.e. relatively close to the orbital phase where the highest x - ray activity of the source has been observed in several occasions .
unfortunately , the poor statistics of this /xrt observation prevented an accurate spectral analysis , and thus we could not investigate the origin of this low flux state . in the other two /xrt observations a spectral analysis could be carried out , but we did not detect any indication of a significant spectral variation . only in the observation we measured a slight increase in the spectral power law index .
this suggests that x - ray flux changes in 18483 might have occurred due to genuine variations in the mass accretion rate onto the ns , rather than eclipse - like events .
note that , the detection of pulsations in the data are also in agreement with the accretion scenario .
this suggests that sfxts undergo low level accretion even when they are not in outburst ( see also * ? ? ?
at odds with the case of 18483 , table [ tab : comparison ] shows that the different orbital phases of 18 have been fairly well monitored .
unfortunately , the x - ray spectrum of this source could be well characterized only during the outburst , whereas in quiescence only the observation provided a measurement of the spectral parameters ( see table [ tab : log ] ) . in all the other observations
only a rough estimate of the source flux could be obtained .
note that the source was not detected by at the orbital phase 0.52 , and the 3@xmath33 upper limit on the source x - ray flux was at least an order of magnitude lower than the fluxes measured in any other orbital phases .
since no spectral analysis could be carried out on 18 at this orbital phase , the origin of this low flux event could not be investigated further . in case
future observations of 18 reveal that this source regularly undergoes x - ray eclipses at the orbital phase @xmath40.5 , this can help clarifying the issue of the extreme flux changes in this source .
more observations of 18483 and 18 at different orbital phases with high sensitivity x - ray telescopes , like and , are clearly required in order to understand unambiguously the origin of their outburst / quiescent activity .
being these two sources the only sfxts with a comparable orbital period , they are very well suited to test different models proposed to explain the behavior of sfxts .
we are currently investigating the results of the application of the gated accretion model to 18483 and 18 ( bozzo et al .
, 2009 , in preparation ) . in this paper , besides x - ray flux changes , we also measured a spin - period variation in 18483 . by using our best - determined spin - period , @[email protected] s , and that found previously by @xcite
, we obtained a spin - period derivative in 18483 of @xmath73=-([email protected])@xmath110@xmath2 s s@xmath3 .
this value is comparable with the spin - period derivative measured in the case of the sfxt axj1841.0 - 0535 ( -1.5@xmath110@xmath57
s / s * ? ? ?
* ) and those induced by accretion torques in wind - fed binaries ( see e.g. , * ? ? ?
* ) . however , in the present case we believe that the spin - period derivative most likely results from light travel time effects in the binary .
in fact , in a binary system with an orbital period of @xmath418.5 days , these effects can contribute to an apparent spin - period derivative of the order of @xmath4@xmath74/@xmath75=8.6@xmath110@xmath76 s s@xmath3 , i.e. much larger than the spin - period derivative we reported above ( here @xmath74 is the orbital velocity and @xmath75 is the light velocity ; we used the mass and radius of the supergiant companion measured by * ? ? ?
unfortunately , since a detailed orbital solution for this source is not yet available , we do not know if accretion torques acting onto the ns in 18483 might also have contributed to the observed spin - period derivative .
note that , in principle , this can be used to study the interaction between the ns and the inflowing matter from the supergiant companion ( see e.g. * ? ? ?
an orbital monitoring of 18483 is required in order to understand the origin of the measured spin - period derivative .
we thank the anonymous referee for his / her many helpful comments .
eb thank m. capalbi , m. perri , and k. page for kind help with the /xrt data analysis .
this work was partially supported through asi and miur grants . | igrj18483 - 0311 was discovered with in 2003 and later classified as a supergiant fast x - ray transient .
it was observed in outburst many times , but its quiescent state is still poorly known . here
we present the results of , , and observations of 18483 .
these data improved the x - ray position of the source , and provided new information on the timing and spectral properties of 18483 in quiescence .
we report the detection of pulsations in the quiescent x - ray emission of this source , and give for the first time a measurement of the spin - period derivative of this source . in 18483 the measured spin - period derivative of -([email protected])@xmath110@xmath2 s s@xmath3 likely results from light travel time effects in the binary .
we compare the most recent observational results of 18483 and 18 , the two supergiant fast x - ray transients for which a similar orbital period has been measured .
x - rays : binaries - stars : individual ( 18483 , 18 ) -stars : neutron - x - rays : stars |
the development of experimental and theoretical methods of quantum information processing ( qip ) is an important direction of quantum informatics .
nmr methods are the simplest ones among different experimental approaches @xcite .
it is well known that the liquid state nmr methods @xcite are widely used for experimental realizations of quantum gates and algorithms on the basis of the pseudo - pure states @xcite
. however , the possibilities of qip based on the liquid state nmr are limited because the number of correlated qubits is very small ( about 10 ) @xcite and quantum correlations in such systems are not strong .
in particular , entanglement is almost absent @xcite . at the same time , the potential of the solid state nmr methods @xcite is not exhausted yet .
multiple quantum ( mq ) spectroscopy in solids @xcite is an example of such a method .
mq nmr not only creates multi - qubit coherent states but also allows the investigation of their relaxation under the action of the correlated spin reservoir .
mq nmr @xcite is an important method for the investigation of various problems of quantum information processing such as the transmission of quantum information @xcite and decoherence processes @xcite .
one - dimensional mq nmr methods are very suitable for solving problems of quantum informatics .
the reason is that a consistent quantum - mechanical theory of mq nmr dynamics has been developed only for one - dimensional systems @xcite .
that theory is based on the fact that the non - secular two - spin / two - quantum hamiltonian @xcite , describing mq nmr dynamics , is the xy hamiltonian @xcite , which can be diagonalized exactly for one - dimensional systems in the approximation of the nearest neighbor interactions @xcite . as a result , mq nmr dynamics in such systems can be studied analytically . in particular , only mq nmr coherences of the zeroth and plus / minus second orders arise in a one - dimensional chain initially prepared in a thermodynamic equilibrium state @xcite .
the developed theory @xcite is based on the model of an isolated spin chain with the nearest neighbor interactions .
such a model is realized in quasi - one - dimensional chains of nuclei in calcium fluorapatite with a hexagonal system of fluorine nuclei .
the distance between neighboring chains is about three times larger than the distance between the nearest fluorine nuclei in the same chain .
taking into account that the dipole - dipole interaction ( ddi ) of the next nearest spins in the chain is eight times weaker than the nearest neighbor interaction @xcite one can use the model of an isolated spin chain with the nearest neighbor interactions for the interpretation of the experimental data obtained from the mq nmr experiments @xcite .
the information resource of mq nmr exceeds the resource of conventional nmr @xcite .
however , that resource can be used only if the xy hamiltonian is created with sufficient accuracy @xcite . the xy hamiltonian is obtained as the averaged two - spin / two - quantum hamiltonian @xcite which is the average hamiltonian @xcite in the mq nmr experiments .
the corrections to the average hamiltonian @xcite spoil the xy hamiltonian .
these corrections are proportional to powers of the parameter @xmath0 ( @xmath1 is the period of the irradiating sequence of the preparation period of the mq nmr experiment and @xmath2 is of the order of the ddi ) .
we have designed a special probe @xcite for the generation of high - power ultra - short pulses and decreased the period @xmath1 and the parameter @xmath3 in mq nmr experiments .
the results @xcite show high accuracy of creating the xy hamiltonian .
it is very important that the xy hamiltonian ( the mq nmr hamiltonian ) is related to the flip - flop one by a simple unitary transformation @xcite for one - dimensional systems in the approximation of the nearest neighbor interactions .
that fact allows us to realize the quantum state transfer along a one - dimensional spin chain .
it is also significant that the relaxation of the mq nmr coherences can be considered as a model for studying the decoherence processes .
the paper is organized as follows .
an introduction to mq nmr dynamics is given in section 2 . in section 3
we show that mq nmr can be used for the quantum state transfer in one - dimensional spin chains .
relaxation of the mq nmr coherences in one - dimensional spin chains is discussed in section 4 .
we briefly summarize our results in section 5 .
[ cols="^ " , ]
we considered the methods of mq nmr spectroscopy of one - dimensional spin systems and their application for solving problems of quantum informatics . the quantum state transfer in linear spin chains
can be realized with high fidelity using mq nmr dynamics .
decoherence of many - qubit quantum state created on the preparation period of the mq nmr experiment @xcite can be studied on the evolution period of that experiment . the dipolar relaxation of the mq nmr coherences is considered as a model of the decoherence process .
the theory of the dipolar relaxation of the mq nmr coherences is developed .
the performed mq nmr experiments on the quasi - one - dimensional system of nuclei in calcium fluorapatite are in a good agreement with the theoretical predictions .
braunstein , s. l. , caves , c. m. , jozsa , r. , linden , n. , popescu , s. , and schack , r. , `` separability of very noisy mixed states and implications for nmr quantum computing , '' _ phys .
lett _ * 83 * , 1054 ( 1999 ) .
doronin , s. i. , vasilev , s. g. , samoilenko , a. a. , feldman , e. b. , and shumm , b. a. , `` dynamics and relaxation of multiple quantum nmr coherences in a quasi - one - dimensional chain of nuclear spins @xmath4f in calcium fluorapatite , '' _ jetp letters _
* 101 * , 613 ( 2015 ) .
bochkin , g. a. , feldman , e. b. , and vasilev , s. g. , `` relaxation of multiple quantum nmr coherences in quasi - one - dimensional spin systems , '' _ zeitschrift fr physikalische chemie international journal of research in physical chemistry chemical physics _
* 230 * ( 2016 ) . | multiple quantum ( mq ) nmr methods @xcite are applied to the analysis of various problems of quantum information processing .
it is shown that the two - spin / two - quantum hamiltonian @xcite describing mq nmr dynamics is related to the flip - flop hamiltonian of a one - dimensional spin system in the approximation of the nearest neighbor interactions . as a result , it is possible to organize quantum state transfer along a linear chain .
mq nmr experiments are performed on quasi - one - dimensional chains of nuclei in calcium fluorapatite .
relaxation of the mq nmr coherences is considered as the simplest model of decoherence processes .
a theory of the dipolar relaxation of the mq nmr coherences in one - dimensional systems is developed .
a good agreement of the theoretical predictions and the experimental data is obtained . |
broad band photometry from wide - field imagers mounted on dedicated telescopes and instruments has been and will continue to be our main source of information for a large fraction of the extragalactic universe .
spectroscopy provides a more detailed and deeper understanding of the physical properties of individual objects than photometry .
however , spectroscopy will never be able to fully sample the populations of galactic or extragalactic objects in either number or depth .
it is therefore of great interest to determine whether it is possible to extract estimates of physical parameters , such as distance , metallicity , star formation rate , morphology and the presence or absence of an active nucleus , from the coarse information provided by photometry . for a given object , photometric quantities , such as magnitudes in different bands , momenta of the light distribution and morphological indexes , define its position in a high dimensional parameter space which we shall call the observed parameter space ( or ops , cf .
@xcite ) . the de - projection of the ops into the physical parameter space ( pps , i.e. the parameter space defined by the physical quantities ) , is however a complex operation , made in some cases almost impossible by the degeneracy existing in both the data and the physical parameters themselves .
most of the time , such de - projections require an intermediate passage , i.e. the de - projection of the ops onto the spectroscopic parameter space ( or sps ) , i.e. the space defined by observable spectroscopic quantities such as redshift , equivalent widths of specific spectral lines and spectroscopic indexes .
the parameters defining the sps are in fact more directly related to the intrinsic physical properties of the objects .
spectroscopy , for instance , is more effective than photometry in disentangling normal galaxies from those hosting an agn , or in classifying different types of agn in broad classes ( e.g. , seyfert , liners , etc . ) . actually , the definition of nearly all agn classes is based on spectroscopic criteria through the equivalent widths of some lines ( see for instance @xcite ) .
the problem of mapping the ops onto some subspaces of the sps and then on the pps , is frequently encountered in the literature .
examples include the determination of photometric redshifts , which is an attempt to reproduce what is decidedly a spectroscopic quantity ( the redshift ) using photometry alone ; the search for candidate quasars and blazars using photometric diagnostics ; or the search for candidate agn and starburst galaxies .
usually these mapping functions are quite complex and difficult ( if not plainly impossible ) to be derived in simple analytical forms and require statistical or empirical approaches which are common in the field of machine learning ( hereinafter ml ) .
in fact , if higher accuracy information is available for a subset of objects ( in the case discussed here , spectroscopic parameters or flags in the sps ) , it can be used to teach a ml method how to map the available data ( the photometric data in the ops ) into the higher accuracy ones . in the ml field
this approach is usually called supervised learning .
we emphasize that , while supervised learning is very powerful in uncovering the hidden relationship between input parameters and the so - called knowledge base ( hereafter kb ) , there are two main limitations : * the mapping function can not be extrapolated outside of the region of the ops that is properly sampled by the sps ; * any bias present in the kb is necessarily reproduced ( if not amplified ) in the output . to better exemplify these two points
let us focus on the main topic addressed in this work where we use a specific ml method , the mlpqna @xcite , to tackle a particular incarnation of the so called _ physical classification of galaxies _ problem .
we will specifically address the possibility of disentangling normal , non - active galaxies from those hosting an agn using only photometric measurements . in a previous paper
we have discussed the use of the same method to classify different types of agn @xcite . in this work
we shall instead focus on the question of whether it is possible to use only optical colors to produce reliable ( or at least statistically well controlled ) agn candidates and to disentangle normal - inactive galaxies from those hosting an agn . as already pointed out in @xcite , in spite of the unique physical mechanism responsible for the nuclear activity @xcite , the phenomenological complexity of agns is so high that there can not be a unique method equally effective in identifying all agn phenomenologies in every redshift range ( cf .
moreover , agn and starbursts , long studied separately , are now thought to be correlated @xcite and difficult to disentangle without using mid - ir data .
furthermore , we also investigate the possibility of identifying a catalogue of candidate qsos , by comparing the efficiency of classification in terms of a compromise between purity and completeness following two different strategies : i ) by exploiting a self - consistent strategy starting from three different classes of objects ( for instance galaxies , stars and qsos ) ; ii ) by assuming a pre - determined separation between resolved and un - resolved objects ( star / galaxy ) , we performed the usual two - class classification between stars and qsos .
the paper is structured as follows . in section [ sec : thedata ] we discuss in same detail the data . in section
[ sec : themethod ] we summarize the general methodology adopted and provide a short description of the ml method used . in section [ sec :
experiments ] we describe in some detail the various classification experiments performed and , finally , in section [ sec : conclusions ] we discuss the results .
among the countless merits of the sloan digital sky survey @xcite , there is also the fact that it has paved the way to the experimenting and wide adoption within the astronomical community of innovative methods based on ml ( or on statistical pattern recognition ) , thus fostering the birth of the emerging field of astroinformatics @xcite .
the sdss in fact provides an ideal test ground for ml algorithms and methods : a large and complex photometric database with hundreds of features measured for hundreds of millions of objects ( defining the sdss subregion of the ops ) , complemented by spectroscopic information for a significant subsample ( roughly @xmath3 ) of objects ( i.e. the sps ) .
furthermore , it offered to different groups , working with different methodologies , the possibility to address significant problems , thus allowing a robust and fair assessment of the performance of any new method . even constraining ourselves to the specific interest of the present work
, it would be difficult to summarize the applications of ml methods to sdss data and we shall just mention two .
the search for candidate quasars has been widely discussed in @xcite . while the characterization of agn has been discussed in @xcite .
last but not least , ml methods have been crucial for the first successful astronomical citizen science project : the galaxy zoo @xcite , where empirical methods were used to compare and evaluate the performance of a large number of human simple , repetitive , and independent classification tasks . in this paper
we use photometric and spectroscopic data extracted from the sdss data release 10 ( dr10 ; @xcite ) .
the dr10 photometry covers @xmath4 @xmath5 of the celestial sphere for a total of more than @xmath6 millions unique ( i.e. without duplicates , overlaps and repeated measurements ) objects , but not necessarily unique _
astrophysical _ objects .
dr10 also includes spectroscopic information for more than @xmath7 million objects .
these data come from a wide range of concurrent experiments : the @xmath8 initial sdss data releases , the sloan extension for galactic understanding and exploration ( segue-2 , @xcite ) , the apache point observatory galaxy evolution experiment ( apogee ; @xcite ) mainly targeted to milky way stars ; the barion oscillation spectroscopic survey ( boss ; @xcite ) and , finally , the multi object apo radial velocity exoplanet large - area survey ( marvels ; @xcite ) . in order to build our kb we used the spectroscopic classification flags provided by the sdss dr10 .
it includes also a spectroscopic classification ( parameters @xmath9 and @xmath10 ) obtained by the sdss team through the comparison of individual spectra with templates and from equivalent width ratios .
this classification is articulated in three main classes : @xmath11 , @xmath12 and @xmath13 as follows : + * _ galaxy_*. objects having a galaxy template .
its subclasses are : * starforming : based on whether the galaxy has detectable emission lines that are consistent with star - formation according to the criteria : @xmath14 * starburst : set if the galaxy is star - forming but has an equivalent width of @xmath15 greater than @xmath16 ; * agn : based on whether the galaxy has detectable emission lines that are consistent with being a seyfert or liner according to the relation : @xmath17 * _ qso_*. objects identified with a qso template . based on the sdss definition ,
if a galaxy or quasar has spectral lines detected at the @xmath18 sigma level with @xmath19 at the @xmath20 sigma level , the indication _ broadline _ is appended to their subclass . + * _ star_*. objects matching a stellar template ( among @xmath21 spectral subclasses ) .
since we were not interested in these subclasses , they will be ignored in what follows .
+ the total number of spectroscopic objects is @xmath22 divided in different types as summarized in table [ tab : kbdetails ] .
we assumed all galaxies not flagged otherwise to be normal , i.e. neither agn or starburst .
it is worth noting that sharp cuts based on crisp thresholds introduce ambiguities in the mapping of the sps onto the pps . to be more clear : a large fraction of the objects with spectroscopic features near the threshold values
may be misclassified ( i.e a galaxy hosting an agn may be erroneously put in the normal galaxy bins due spectroscopic errors and viceversa ) .
these ambiguities are intrinsic to the kb and will affect the projection of the ops onto the sps and therefore the pps
. for all spectroscopic objects we downloaded the sets of magnitudes listed in the upper part of table [ tab : features ] .
the choice of using different sets of magnitudes was dictated by the fact that , since we were interested also in finding agn candidates , different apertures can be used to weight in different ways the contribution of the central unresolved source and of the extended surrounding galaxy . [ cols="<,<,>,>",options="header " , ] the results of the @xmath23 experiments are reported in table [ tab : threeclass ] . by looking at the qso statistics
, the purity / completeness results of @xmath24 experiment appeared very similar to those obtained in the @xmath25 experiment .
although these two experiments led to comparable results , we consider the @xmath24 as the best , because obtained without any cut in the parameters . for a complete discussion
[ sec : conclusions_1 ] .
the classification experiments described in the previous sections aimed at producing a reliable catalogue of candidate qsos / agns by exploiting the photometric information available within the dr10 . in what follows we give a detailed overview of their results .
the set of experiments described in sec .
[ sec : normalgal ] aimed at investigating whether optical photometry could be used to isolate different types of galaxies as defined by the spectroscopic subclasses in the sdss dr10 spectroscopic archive . with reference to table [ tab : kbdetails ] it is apparent that different types of objects are represented in the kb in a very uneven way . in the first couple of experiments ( _ all _ and _ cuts _ ) we tried to separate normal galaxies from those having a _ peculiar _ spectrum ( i.e. all the other subclasses in the sdss classification scheme ) .
the performance of the method , shown in table [ tab : goresults ] , were not affected by the application of color and magnitude cuts and the efficiency remained stationary around a value of @xmath26 . given the lack of improvements between the _ all _ and _ cuts _ experiments , we decided to perform supplementary experiments by freezing the _ all _ kb , the same configuration for the network and the input parameters .
the statistics of misclassified objects , reported in tables [ tab : misall ] and [ tab : miscuts ] , show that the mlpqna is quite effective in disentangling starburst and starforming galaxies both narrow and broadline ( misclassification rates of @xmath27 and @xmath28 , respectively ) , and much less effective in identifying other types of objects such as , for instance , agn ( misclassification rate @xmath29 ) and agn broadline objects ( @xmath30 ) .
the performance on the starburst broadline and starforming broadline types , however , are difficult to evaluate , since these two groups are very poorly represented in the kb and therefore the network may have failed in capture their specific signatures in the ops .
the fact that @xmath29 of the agn , @xmath31 of the broadline and @xmath32 of the agn broadline were misclassified induced us to check whether different groupings of the subclasses could lead to improvements in the classification efficiency . in the _
nbonly _ experiment ( table [ tab : nbonly ] ) the misclassification rate improved to @xmath33 for the normal galaxies and dropped to @xmath33 ( from @xmath31 ) for the broadline .
the misclassification rates for agn broadline , starburst broadline and starforming broadline increased , confirming the fact that the network is not capable to disentangle these spectroscopic subclasses in an effective way . in the _ nball _ experiment ( table [ tab : nball ] )
the results did not change in a significant way .
in fact , the normal , starburst and starforming type objects still maintain a misclassification rate of , respectively , @xmath33 , @xmath27 and @xmath28 .
while for the broadline galaxies the misclassification decreases to @xmath28 .
a slightly better result is obtained for the agn broadline type , going from @xmath32 in _ all _ and @xmath31 in _ nbonly _ , to @xmath34 .
the agn and starforming broadline types are even worse reaching a misclassification of , respectively @xmath35 and @xmath36 . finally , the starburst broadline , although sparsely represented in the data set , resulted almost completely misclassified ( @xmath37 ) .
the last experiment ( _ nba _ ) , resulted as the best one in terms of overall efficiency as well as for classes normal , broadline , agn broadline and starburst broadline , where the misclassification rates decreased to , respectively , @xmath38 , @xmath39 , @xmath40 and @xmath33 .
only starburst type rate remains unchanged ( @xmath27 ) , while agn objects perform slightly worse ( @xmath41 vs @xmath29 obtained in the _ all _ case ) . on the other hand , the classification of starforming and starforming broadline type decreases to , respectively , @xmath33 and @xmath42 , although the starforming rate decreases by @xmath27 only . the strong decrease for
the starforming broadline type could be due to its poor presence within the data set .
already in the earliest sdss works @xcite optical colors were used to disentangle stars from quasars in the sdss color space .
@xcite used an unsupervised clustering algorithm followed by an agglomeration phase to identify candidate qsos in a parameter space based only on photometric colors .
@xcite used a difference boosting neural network ( dbnn ) on sdss optical colors only , achieving excellent performance ( @xmath43 ) in disentangling qso from normal galaxies in the unresolved source catalogues ( thus including stars , quasars and unresolved galaxies ) .
they however applied a cut in colors by isolating the region of this space where most ( @xmath37 ) of the spectroscopically confirmed quasars lay .
this choice , however , penalizes the recognition of peculiar objects which are likely to lay outside of the main distribution of normal _ well behaved _ objects .
our experiments are described in sec .
[ sec : twoclass ] and related results are shown in table [ tab : twoclass ] . the immediate conclusion which can be drawn by analyzing the results is that our method is able to achieve @xmath1 of completeness and a purity of @xmath44 without having to apply any color cuts .
in the three - class classification experiment the kb included all objects in the spectroscopic catalogue regardless their extended or unresolved nature .
the first thing to notice is that all experiments reported in table [ tab : threeclass ] have performance which differ only by very little amount .
the experiment which led to the best results was the @xmath24 with an overall efficiency of @xmath45 .
this experiment was performed using a @xmath46 layers mlpqna and a quite small training set ( only @xmath47 of the available kb ) using as input parameters the @xmath18 sdss magnitudes .
having identified the experiment @xmath24 as the best one , two additional experiments were performed on the @xmath24 parameter space , by changing either the topology of the neural network or the amount of training set to evaluate their individual contribution to the performance . in the first case we used a @xmath48 layer mlpqna and in a second case a larger training set ( @xmath49 of the kb ) .
the case with the increased training set led to a negligible improvement in the overall efficiency but at the price of a huge increase in computing time . while , in the case of a single layer mlpqna network , the performance decreased of about @xmath50 .
this can be understood by taking into account that the complexity introduced by the second hidden layer in the architecture of the neural network is always justified when the kb samples the ops only sparsely . in the case
discussed here , even with a reduced kb , the number of samples in the training set was sufficiently high to ensure a proper coverage of the ops and to minimize the requirements on generalization capabilities .
the additional complexity introduced by the second layer proves indeed very useful in presence of much smaller kbs @xcite .
surprisingly enough , the introduction of mag @xmath51 cuts in the parameter space ( experiment @xmath25 ) did not lead to any significant improvement in the method efficiency . in summary
, using optical data only , the mlpqna seems quite effective in identifying galaxies ( completeness @xmath52 and contamination @xmath53 ) from qso and stars ( table [ tab : threeclass ] ) .
we performed the supervised learning classification experiment to produce a catalogue of high fidelity candidate qso selected on photometric data only .
the classification is based on a membership probability criterium : for each object the model output is based on three different membership pseudo - probabilities ( a confidence level for each class ) . therefore the values in table [ tab : threeclass ] are referred to these confidence levels for all objects of the blind test set ( technique also known as _ winner - takes - all _ ; @xcite ) . with this method ,
in the case of the selected best experiment , we reached a qso class purity of @xmath54 .
afterwards , since our goal was to reach the highest level of purity in the produced catalogue , we performed a further statistical analysis of the test set starting from the results of the experiment @xmath24 , by assessing the variation of purity vs completeness as a function of the increasing confidence threshold used to evaluate the qso candidates from the trained mlpqna model output . at the end we reached a purity of @xmath1 in the blind test set , at the price of a reduced completeness .
the resulting qso photometric catalogue contains @xmath55 candidates and will be made publicly available through the cds vizier facility .
we wish to emphasize that a three - class classification approach offers a significant advantage with respect to the traditional approach based on serialized series of two - class classification steps , since it does not introduce systematic biases due to objects which are misclassified at each step of the sequence . specifically : the separation among star , galaxies and quasars could be achieved via two independent classification steps : first separating resolved ( galaxies ) from unresolved objects ( stars and quasars ) and then dividing the latter group into stars and quasars .
objects misclassified at the first step would be propagated into the next step . restricting ourselves to the qso vs star classification , in the three - class experiment we achieve an overall accuracy of @xmath56 , with a qso purity of @xmath57 and qso completeness of @xmath58 .
therefore , by comparing this result with those of the two class experiment ( table [ tab : twoclass ] ) there is a noticeable improvement of accuracy in the three - class experiment .
this confirms the fact that a three - class approach is preferable to the hierarchical chain of two - class experiments .
the slight decrease in qso completeness found in the three - class experiment may be simply due to the higher separation complexity of the experiment .
it needs to be stressed , however , that the catalogue of candidate quasars needs to be used with some cautions .
in fact , no a - priori defined photometric cuts have been applied to the data and our operative definition of a quasar was induced only by the properties of the spectroscopic knowledge base .
hence , any bias , potentially present in the kb , would be reflected in the final definition of the catalogue .
let us , for instance , take into account the bright end of the luminosity distribution .
it is well known that the sdss quasar catalogue is fairly ( @xmath59 ) complete for @xmath60 .
if we consider the produced catalogue , within the flux limit of @xmath61 we find @xmath62 more candidates than those present in the spectroscopic sample .
but , as already discussed in the literature @xcite , this bright end is strongly contaminated by stars , uvx sources and , mainly , by narrow emission - line galaxies . in order to isolate luminous quasars , @xcite used a complex system of photometric cuts : a ) : : u - g@xmath63 g - r@xmath64 r - i@xmath65 ; b ) : : u - g@xmath66 g - i@xmath67 ; c ) : : u - g@xmath68 g - i@xmath69 ; d ) : : galprob@xmath70 u - g@xmath71 g - r@xmath72 r - i@xmath73 ; e ) : : galprob@xmath70 u - g@xmath74 .
these cuts were indeed combined through the logical expression [ @xmath75 , to carve the optimal quasar locus in the observed parameter space .
if we apply the above conditions to both spectroscopic kb and the final catalogue of candidate quasars , under the flux limit of @xmath61 the overabundance of objects in the produced catalogue is reduced of the @xmath76 . in other words , while ml methods prove very effective in partitioning the ops accordingly to the information contained in the spectroscopic kb , the correct interpretation of their output needs to be fine tuned using the expert s ( i.e. the astronomer s ) knowledge . for this reason , in order to allow interested readers to apply their own filters to the data , the catalogue contains all the relevant photometric information . in fig .
[ fig : histo1 ] we compare the number of candidate qsos ( solid black line ) in our catalogue vs the number of objects in the kb which are spectroscopically identified as qsos ( solid grey line ) , both before and after ( dashed lines ) the cuts of @xcite . before the cuts ,
the excess of spectroscopically confirmed agns / qsos at bright magnitudes ( @xmath77 ) can be easily explained by the fact that low luminosity agns , which are spectroscopically identified , can easily escape photometric detection since the relative weight of the agn contribution is negligible with respect to the contribution of the central regions of the galaxy .
furthermore , in this magnitude range the sdss sample is highly unbalanced , thus causing the mlpqna model to be less accurate in terms of training performance on the qso class . after applying the @xcite selection criteria , however , the agreement between the two distributions becomes remarkably good and it is maintained down to the completeness limit of the sdss spectroscopic sample ( @xmath78 ) . beyond such limit , the two distributions begin to differ due to both new candidate quasars which were not spectroscopically confirmed in the sdss ( region between the two dashed lines ) and to a contamination from low / medium luminosity agns ( region between the two black lines ) which , as it has been discussed in sec .
[ sec : galvsothe ] , can not be disentangled on the grounds of optical photometry only .
the trends depicted in the diagram of fig .
[ fig : histo1 ] find a statistical confirmation in fig .
[ fig : cdf4 ] , where we show the results of the 1d kolmogorov - smirnov ( k - s ) test @xcite , graphically represented through the cumulative @xmath79 band distribution functions .
the test has been performed to compare the distributions of qso / agn candidate sources , respectively , between the photometric catalog and the spectroscopic kb , before and after the completeness limit of the sdss spectroscopic sample ( @xmath78 ) and the @xcite cuts application . within the completeness limit
, the k - s test confirms the similarity of the two distributions , perfectly overlapping after the application of @xcite cuts . beyond the completeness limit
, the k - s test makes evident a difference between the two distributions , considered reasonable by taking into account the incompleteness of the spectroscopic sample . in the produced catalogue ,
a quality flag has been included to take the @xcite cuts into consideration .
the main scope of the present work was to investigate the possibility of disentangling different spectral classes of objects , by exploiting the machine learning based model named mlpqna @xcite . in order to reach these goals , three categories of experiments
have been performed . in a first group of experiments we investigated the possibility to disentangle different spectroscopic types present in the galaxy class .
all experiments , summarized in the tables [ tab : goresults ] and [ tab : nb ] were unable to separate agns from normal galaxies on the grounds of optical data only , although the overall efficiency is always @xmath26 , thus confirming what we found in @xcite . by excluding the agn class , in all other cases where the number of representative objects in the kb is sufficiently large , the disentangling appears quite feasible , in particular by considering results of the experiment _ nba _ ( table [ tab : nba ] ) .
the second one concerned distinguishing between qso and star classes .
the results , shown in table [ tab : twoclass ] , confirmed the findings of @xcite , that the use of @xmath18 colors , rather than the standard @xmath80 colors , although without carrying a great amount of additional information , may help machine learning methods to find better solutions .
the third category focused on identifying candidate qsos from the whole catalogue including also stars and galaxies , thus permitting to release the candidate qso photometric catalogue . in this case , the possibility of avoiding the known downside of hierarchical pairwise classification ( multiplicative propagation of misclassification ) , induced an intrinsic advantage of a direct classification of qsos from the whole catalogue .
the resulting catalogue of qso candidates contains @xmath55 objects , of which @xmath81 are flagged as robust candidates , according to the quality flag introduced to take into account the @xcite selection criteria .
this catalogue will be made publicly available through the cds vizier facility .
as discussed in the text , the catalogue requires to be filtered according to the specific needs of the user .
the authors wish to thank the anonymous referee for all very useful comments and suggestions which greatly helped to clarify and to improve the readability of the paper .
the authors wish to acknowledge the financial support from the prin 2011 miur grant _ cosmology with euclid_. mb acknowledges financial support from prin - inaf 2014 _ glittering kaleidoscopes in the sky , the multifaceted nature and role of galaxy clusters_. we made use of topcat tool developed within the virtual observatory , and the data mining infrastructure dameware . this research has made use of the sdss iii dr10 and vizier catalogue data access tools .
djorgovski s.g .
, mahabal a. , drake a. , graham m. , & donalek c. 2012 , in astronomical techniques , software , and data ( ed . h. bond ) , vol.2 of planets , stars , and stellar systems ( ser .
t. oswalt ) , p. 223 .
berlin : springer verlag | we discuss whether modern machine learning methods can be used to characterize the physical nature of the large number of objects sampled by the modern multi - band digital surveys .
in particular , we applied the mlpqna ( multi layer perceptron with quasi newton algorithm ) method to the optical data of the sloan digital sky survey - data release 10 , investigating whether photometric data alone suffice to disentangle different classes of objects as they are defined in the sdss spectroscopic classification .
we discuss three groups of classification problems : _
( i ) _ the simultaneous classification of galaxies , quasars and stars ; _ ( ii ) _ the separation of stars from quasars ; _ ( iii ) _ the separation of galaxies with normal spectral energy distribution from those with peculiar spectra , such as starburst or starforming galaxies and agn . while confirming the difficulty of disentangling agn from normal galaxies on a photometric basis only , mlpqna proved to be quite effective in the three - class separation . in disentangling quasars from stars and galaxies , our method achieved an overall efficiency of @xmath0 and a qso class purity of @xmath1 .
the resulting catalogue of candidate quasars / agns consists of @xmath2 million objects , of which about half a million are also flagged as robust candidates , and will be made available on cds vizier facility .
[ firstpage ] methods : data analysis - techniques : photometric - catalogues - galaxies : active - quasars : general |
consider a quantum system on an infinite lattice described by the hamiltonian @xmath0 , with @xmath1 a dimensionless coupling constant .
for any reasonable @xmath1 , all observable properties of the _ ground state _ of @xmath2 will vary smoothly as @xmath1 is varied
. however , there may be special points , like @xmath3 , where there is a non - analyticity in some property of the ground state : we identify @xmath4 as the position of a quantum phase transition . in finite lattices ,
non - analyticities can only occur at level crossings ; the possibilities in infinite systems are richer as avoided level crossings can become sharp in the thermodynamic limit . in this paper , i will restrict my discussion to second order quantum transitions , or transitions in which the correlation length and correlation time diverge as @xmath1 approaches @xmath4 .
as i will review below , any such quantum transition can be used to define a continuum quantum field theory ( cqft ) : the cqft has no intrinsic short - distance ( or ultraviolet ) cutoff .
the main purpose of this paper is to review some recent work @xcite on the properties of @xmath5 at _ finite temperatures _ ( @xmath6 ) in the vicinity of @xmath3 .
this is equivalent to a study of the finite @xmath6 crossovers of the associated cqft .
we shall focus especially on the dynamic properties of a ubiquitous finite @xmath6 region , usually called `` quantum critical '' @xcite ( as we shall see below , there are reasons why this name is misleading and not quite appropriate ; nevertheless , i will use it here ) .
the quantum - critical region appears as the high @xmath6 limit of the cqft ; unlike the statics of classical lattice models , the high @xmath6 limit of a cqft is usually highly non - trivial .
all of this discussion will take place in the context of some simple examples drawn from quantum spin systems .
i set the stage by reviewing the wilsonian approach to critical phenomena and field theories @xcite , using the perspective of quantum critical phenomena . by the usual trotter product decomposition
, we can set up the partition function of @xmath0 as a functional integral over degrees of freedom which fluctuate as a function of the spatial co - ordinate @xmath7 and imaginary time @xmath8 .
let us now examine the behavior of this functional integral under the rescaling transformation @xcite @xmath9 the dynamic exponent @xmath10 determines the relative scaling dimensions of space and time co - ordinates .
the critical point at @xmath3 is a fixed point of ( [ rg ] ) , and @xmath11 is a relevant perturbation away from this point .
we have therefore the flow equation @xmath12 which defines the critical exponent @xmath13 .
( for simplicity i do not discuss the case of fixed points with more than one relevant perturbation , as they can be treated in a similar manner ) . in the long - distance , long - time limit
, this deviation from the critical point will be characterized by some _
renormalized _ energy scale , @xmath14 .
i emphasize that @xmath14 is a dimensionful parameter , expressed in the laboratory units of energy , and directly measurable in an experiment ; a typical example would be an energy gap . in the vicinity of the critical point
, the renormalized energy scale @xmath14 will be related to the bare coupling @xmath1 by @xmath15 where @xmath16 is an ultraviolet cutoff , measured for convenience in the units of energy too . from the perspective of a field theorist ,
the cqft associated with the quantum critical point is now defined by taking the limit @xmath17 at fixed @xmath14 ; from ( [ gg ] ) we see that , because @xmath18 , it is possible to take this limit by tuning the bare coupling @xmath1 closer and closer to the critical point as @xmath16 increases .
( a condensed matter physicist would take the complementary , but equivalent , perspective of keeping @xmath16 fixed but moving closer to criticality by lowering his probe frequency @xmath19 ) .
the resulting cqft then contains only the energy scale @xmath14 . at finite temperatures ,
there is a second energy scale @xmath6 ( using units in which @xmath20 ) ; its thermodynamic properties will then be a universal function of the only dimensionless ratio available@xmath21 .
it is the purpose of this paper to review recent work on the crossovers as a function of @xmath21 in a number of systems , and to highlight the unusual properties of the heretofore unexplored high - temperature , `` quantum critical '' , limit of the cqft , @xmath22 .
it is now easy to see why the high @xmath6 limit of the cqft can be non - trivial . a conventional high @xmath6 expansions of the lattice model @xmath2 proceeds with the series @xmath23 the successive terms in this series are well - defined and finite because of the ultraviolet cutoffs provided by the lattice .
further , the series is well - behaved provided @xmath6 is larger than all other energy scales ; in particular we need @xmath24 .
in contrast , the cqft was defined by the limit @xmath17 at fixed @xmath6 , @xmath14 , so the high @xmath6 limit of the cqft corresponds to the intermediate temperature range @xmath25 of the lattice model .
it is not possible to access this temperature range by an expansion as simple as ( [ hight ] ) , and more sophisticated techniques , to be discussed here , are necessary .
( one could also , of course , determine a large number of terms in ( [ hight ] ) and then use some pad extrapolation methods to access the @xmath26 region : this method has been used by sokol _ et .
@xcite and i will not discuss it here ) .
in contrast to the static properties , the dynamic properties of @xmath2 are already non - trivial in the high @xmath24 limit of the lattice model .
although one expects some sort of incoherent , dissipative dynamics , the damping co - efficients can not be determined directly all the approaches used so far are essentially variants of the methods discussed by moriya @xcite and forster @xcite , and use a short - time expansion , coupled with an ansatz for the spectral function , to extrapolate to the long - time limit . in this paper , we will discuss the dynamics in the high @xmath6 limit of the cqft , @xmath27 , or the `` quantum - critical dynamics '' .
the dynamics continues to be dissipative and relaxational , but is not amenable to a description either by a classical boltzmann equation for a dilute gas of quasiparticle excitations , or by a classical langevin - like models of the types discussed in the classic review of hohenberg and halperin @xcite .
however , as we shall show , the scaling structure of the cqft does permit some progress to be made ; indeed we will discuss below the complete solution of the quantum - critical dynamics of a simple spin model , including the exact determination of a damping co - efficient .
we will begin our discussion with a simple solvable model of spinless fermions in section [ spinferm ] : this will allow introduction of the main concepts in a very simple setting .
section [ sec : bose ] will extend these results to a related but more complex model of dilute bosons in spatial dimensions @xmath28 .
an explicit solution of the relaxational dynamics of the quantum - critical region will be obtained in the discussion in section [ isingsec ] on the ising model in a transverse field .
the expository analyses of these toy models will lead in section [ secrotor ] to the main system of interest the @xmath29 quantum rotor model in @xmath30 .
we will also review applications of these results to numerical simulations and experiments .
finally section [ concsec ] will conclude by highlighting the main results and noting recent work on related subjects .
much of the physics i wish to discuss is displayed in a surprisingly simple model of a dilute gas of spinless fermions at finite temperature : its scaling forms have a structure identical to those of much more complicated models .
the main shortcoming of the model is that the associated cqft has no interactions , and there is therefore no relaxational behavior in the scaling functions .
consider the following hamiltonian : @xmath31 where @xmath32 is a spinless fermion annihilation operator at the site @xmath33 of a @xmath34-dimensional hypercubic lattice , and @xmath35 are nearest neighbors .
there are no interactions in @xmath36 , so it is trivially solvable . consider the ground state of @xmath37 as a function of the dimensionless coupling constant @xmath38 . for @xmath39 ,
the ground state has no particles .
there is a non - analytic onset in the density of particles at @xmath3 , signaling a quantum phase transition .
the bandwidth @xmath40 of the fermions plays the role of the upper cutoff in energy ( @xmath16 ) for this transition , and the critical region defining the applicability of a cqft is roughly @xmath41 .
in fact , it is not difficult to determine the exact effective action of the cqft ( in units with @xmath42 ) : @xmath43 where @xmath44 , @xmath45 , and @xmath46 is the lattice spacing . for this action we can identify by the usual methods the exponents @xmath47 and @xmath48 at the @xmath49 critical point ; taking the density @xmath50 as the order parameter
, we get the exponents @xmath51 and @xmath52 .
further , it is also easy to see that all interactions are irrelevant at this critical point in all dimensions @xmath53 ( the least irrelevant interaction term , @xmath54 , becomes relevant only for @xmath55 ) . finally , for the renormalized energy scale measuring the deviation from the critical point , @xmath14 , we may take @xmath56 , the bare chemical potential in @xmath37 .
note that there is no non - universal scale - factor in the relationship between @xmath14 and @xmath57this is a consequence of the triviality of the critical exponents .
more typical models with anomalous exponents will have non - universal scale - factors .
we show in fig [ fermifig ] the phase diagram of @xmath37 as a function of @xmath58 and @xmath6 .
our interest is primarily in the regions within the hatched lines where the cqft ( [ cqft1 ] ) applies . within this region
there are three physically distinct types of behavior ( a , b and c ) : as @xmath59 is trivially solvable , the universal properties of a , b , and c and the crossovers between them can be exactly determined .
let us describe the regions in turn : \(a ) _ activated _ @xmath60 : the fermions are very dilute , with a density @xmath61 .
quantum effects are suppressed and the particles behave classically .
\(b ) _ fermi or luttinger liquid _ @xmath62 : now quantum degeneracy effects are paramount . at @xmath63 ,
the ground state is a fermi liquid ( in @xmath64 , a luttinger liquid ) ; at finite @xmath6 thermal effects lead to a small number of particle and hole excitations near the fermi surface .
\(c ) _ quantum critical _ @xmath65 : unlike a and b , the temperature @xmath6 is the most important energy scale in this region .
we can set @xmath66 here without much damage ( all corrections involve positive powers of @xmath67 ) .
the energy of a typical excitation in this region is of order @xmath6 and as a result , quantum and thermal fluctuations are equally important . the relationships between the regions becomes clearer upon considering an explicit example of a crossover function .
the density of particles @xmath68 obeys the scaling form @xmath69 where the universal scaling function @xmath70 is given by @xmath71 notice that there are no arbitrary scale - factors in ( [ phin ] ) . in the activated region a ( @xmath72 ) , we have @xmath73 , which is exactly the result we would have obtain from the classical maxwell - boltzmann statistics . in the fermi liquid region
b ( @xmath74 ) , @xmath75 , where @xmath76 is the volume of the unit sphere in @xmath34 dimensions ; this is the fully quantum result obtained by filling up the fermi sphere .
most interesting is the quantum critical region c ( @xmath77 where @xmath78 the value of @xmath79 depends upon the details of the fermi distribution function , and not just its forms in the classical and quantum limits : this illustrates our assertion that quantum and classical effects are equally important in region c. it is also interesting to compare the behavior of the density in the universal region c with the true high temperature limit of the lattice model - region d. in other words , we are going vertically upwards in @xmath6 at @xmath66 in fig [ fermifig ] .
it is easy to compute : @xmath80 where @xmath81 is a non - universal constant ( recall that in this model @xmath56 and @xmath82 ) .
notice the difference between the universal high @xmath6 limit of the cqft ( the first result ) and the lattice high @xmath6 limit .
i hope that this example has illustrated the general principle , and in the remainder of the paper i will make no further reference to non - universal regions like d. it will be implicitly assumed that we are working with the universal continuum theory , and we will describe only regions like a , b and c. before closing our discussion on this deceptively simple model , we discuss the scaling form of observables as a function of momentum @xmath83 and frequency @xmath84 . because of the absence of interactions , the single - particle green s function is trivial ; so we discuss the density - density correlator @xmath85 which has a slightly more interesting structure . as the particles are free , @xmath86 is of course given simply by the lindhard function , which can be manipulated into the scaling form @xmath87 where @xmath88 is a universal complex - valued function related to the lindhard function .
notice that , like the scaling form ( [ phin ] ) , there are again no arbitrary scale factors .
further , @xmath88 is well - defined at @xmath89 , where it yields the dynamic susceptibility of the quantum - critical region c. we will see several other examples of scaling forms like ( [ phin ] ) and ( [ phichin ] ) in this paper , but the scaling functions will not be as simple as they are here .
now we consider the same density onset quantum transition considered in sec [ spinferm ] , but for the case of bosons .
the discussion here is drawn from that of sachdev , senthil and shankar @xcite to which the reader is referred for further details .
unlike the case for spinless fermions , it is no longer possible to ignore the interactions between the particles .
we consider the properties of the following continuum model @xmath90 where @xmath91 is a boson annihilation operator , and @xmath92 is a repulsive interaction of range @xmath93 . like @xmath59 , @xmath94 has a quantum phase transition at @xmath95 , and we will discuss its universal properties here . because of the finite range of @xmath96 , the universality only sets in at distances larger than @xmath46 .
a straightforward rg analysis @xcite of the vicinity of the quantum critical point shows that @xmath96 flows into the @xmath97 fixed point for @xmath98 .
it turns out that @xmath96 is actually dangerously irrelevant for @xmath98 : we do not wish to enter into a discussion of such effects here , and so most of our remaining discussion will be restricted to @xmath99 . for @xmath28 ,
@xmath92 flows into a universal fixed point interaction @xmath100 for some @xmath34 dependent constant @xmath101 .
the scaling structure of this fixed point turns out to be very closely related to the non - interacting spinless fermion model of sec [ spinferm ] for the same value of @xmath34 .
all exponents and scaling forms of the boson and fermion models are identical , but the scaling functions themselves are different .
the crossover phase diagram of the @xmath28 dilute bose gas is essentially identical to the fermion phase diagram in fig [ fermifig ] , but the physical interpretation of the phases is somewhat different : \(a ) _ activated _ @xmath60 : this is essentially identical to the fermion case as the particles are dilute and their quantum statistics plays a negligible role .
\(b ) _ incipient superfluid _
@xmath62 : the ground state is now a superfluid ( in @xmath64 a luttinger liquid ) , but classical thermal fluctuations destroy the long range order at any non - zero temperature . nevertheless , the phase coherence length is large and system behaves like a superfluid at short scales .
\(c ) _ quantum critical _ @xmath65 : this is similar to the fermion case in that @xmath6 is the most important energy scale . however , there are now strong interactions among the particles , leading to an incoherent excitation spectrum
. the system does not display characteristics of a superfluid ground state at any length scale , but instead crosses over directly from free particle behavior at short time scales , to dissipative , relaxational dynamics at long time scales .
the single particle green s function @xmath102 has a non - trivial scaling function @xmath103 ( this form actually holds in all three regions a , b , and c ) @xmath104 we have fourier transformed and analytically continued to the retarded green s function at real frequencies .
this scaling form also holds for the spinless fermions of section [ spinferm ] , but the scaling function then is simply the free fermion form @xmath105 where @xmath106 is a positive infinitesimal .
computing @xmath103 , and other scaling functions , for bosonic quantum critical dynamics is not as easy .
one approach @xcite is to expand in powers of the fixed - point interaction @xmath101 which becomes small as @xmath34 approaches 2 from below : this becomes an expansion in @xmath107 . in @xmath64
, it is possible to make more explicit progress .
it has been argued @xcite that now @xmath108 .
the bosons thence become impenetrable , and their quantum mechanics becomes identical to those of free fermions .
hence , in @xmath64 , the quantum critical dynamics of dilute gases of spinless fermions and boson are described by the same cqft , @xmath59 of eqn ( [ cqft1 ] ) . for the bosonic system we have to supplement @xmath59 with the following non - local relationship between the boson and fermion operators ( essentially a continuum jordan - wigner transformation ) @xmath109 we are not home yet , as evaluating correlators of ( [ jw ] ) under ( [ cqft1 ] )
is not easy .
korepin and slavnov @xcite have succeeded in showing how this problem may be reduced to determining the solution and fredholm determinant of a linear fredholm integral equation . at this stage ,
numerical analysis is required , and some scaling functions have been determined to essentially arbitrary accuracy @xcite . finally , an additional comment about the @xmath110 case .
there is now true superfluidity and a finite temperature phase transition to a normal state , all within region b. the remainder of the phase diagram remains the same as in the @xmath111 case .
unlike the dilute fermi and bose gases , the ising model possesses anomalous exponents . yet
it is simple enough in @xmath64 to allow exact computation of a quantum - critical dynamic correlation function .
we will also review , in this section , the very useful and general mapping between the quantum model and an equivalent classical statistical mechanics model ; we will then discuss the crossovers in the phase diagram like fig [ fermifig ] in the context of the classical model . some of of the following discussion is a review of well - known properties of the ising model @xcite ; our main purpose here is to use the explicit solution in @xmath64 to present a physical interpretation which generalizes to other quantum phase transitions .
we will explicitly discuss the following hamiltonian , describing the ising model in a transverse field in @xmath64 ( we will remark briefly on the generalization to higher @xmath34 ) : @xmath112 where @xmath113 in an overall energy scale , @xmath114 is a dimensionless coupling constant , and @xmath115 are pauli matrices on a chain of sites , @xmath33 . consider the ground states of @xmath116 for small and large @xmath1 in turn
. for small @xmath1 , the second term in ( [ hamising ] ) dominates and the spins all align themselves either in the @xmath117 or @xmath118 directions : there is a spontaneous magnetization and spin - reversal symmetry is broken .
on the other hand , for large @xmath1 , the first term in ( [ hamising ] ) prefers a state which is in a superposition of @xmath119 eigenstates , with different sites uncorrelated : the wavefunction looks like @xmath120 .
these two limits are separated by a phase transition at @xmath121 ( in fact , @xmath122 , exactly , because of a self - duality property of ( [ hamising ] ) ) . in this section
, we shall discuss the finite @xmath6 dynamic properties in the vicinity of @xmath3 .
to begin , we recall the well - known fact @xcite that the @xmath63 correlators of @xmath116 are similar to those in the classical , two - dimensional , ising model given by the partition function @xmath123 with @xmath124 where the sites @xmath125 now lie on a square lattice ( say ) .
this classical model has a phase transition at @xmath126 . at the level of the critical continuum theories ,
the mapping between the classical and quantum models become exact @xcite .
there is a simple relationship between the two - dimensional field theory , with classical degrees of freedom , describing @xmath127 in the vicinity of @xmath126 and the one - dimensional quantum field theory describing @xmath116 near @xmath3 : one simply identifies one of the spatial directions of the classical theory as an imaginary time , and then analytically continues the correlators to real time , to obtain observables of the quantum theory .
this mapping leads immediately to some useful information .
as the classical theory is spatially isotropic , the quantum theory has dynamic exponent @xmath128 .
further , in the classical model the @xmath129 correlator behaves like @xmath130 in momentum space @xcite ( @xmath131 is the two - dimensional momentum of the classical model ) ; analytically continuing this to real time , we obtain for the dynamic susceptibility @xmath132 ( this is the @xmath129 correlator of the quantum @xmath116 and @xmath83 is now a one - dimensional spatial momentum ) as a function of @xmath64 momentum @xmath83 and frequency @xmath84 : @xmath133 here @xmath134 and @xmath135 ( an excitation velocity ) are non - universal constants .
we plot @xmath136 in fig [ chiteq0fig ] .
notice that there are no delta functions in the spectral density , indicating the absence of any well - defined quasiparticles .
instead , we have a critical continuum of excitations .
we can also compute the dynamic local susceptibility , @xmath137 a quantity often measured in neutron scattering experiments : @xmath138 this quantity is the density of states of local spin - flip excitations , and has a divergence as @xmath139 .
our discussion so far has been at @xmath63 , and let us turn now to non - zero @xmath6 .
a finite @xmath6 translates into a _
finite size _
@xmath140 for the classical model along imaginary time direction .
periodic boundary conditions are imposed along the finite direction , and so the classical model has the geometry of a cylinder with circumference @xmath140 .
we now discuss the crossovers at finite @xmath6 in the context of the classical model . at @xmath141 ( @xmath63 )
we can characterize the deviations from criticality by the correlation length latexmath:[$\xi \sim energy scale latexmath:[$g=1/\xi \sim display critical correlations characteristic of the point @xmath126 ; it is only at distances larger than @xmath143 that they become sensitive to the value of @xmath144 and display characteristics of the ordered ( for @xmath145 ) or the paramagnetic ( for @xmath146 ) phase .
now consider the effect of a finite @xmath147 ; there are two distinct possibilities : ( _ i _ ) @xmath148 ( for the quantum model , @xmath149 ) : moving from the shortest to largest length scales , the crossover from critical to non - critical ( either ordered or paramagnetic ) behavior still occurs at a scale @xmath150 ; the length scale @xmath147 has little effect at this point .
the effects of @xmath147 only become apparent at larger scales , at which point it is permissible to use an effective model which characterizes the non - critical ground state .
@xmath151 ( for the quantum model , @xmath152 ) : now the short - distance critical fluctuations see a finite size @xmath147 _ before _ they have had a chance to become sensitive to @xmath144 .
these critical fluctuations are quenched by finite size effects in a universal way .
the resulting non - critical theory then responds only weakly at the scale @xmath143 .
note that the system does not display characteristics of the ordered or the paramagnetic state , of the @xmath153 system , at _ any _ length scale .
the above arguments are summarized in figs [ isingfig ] and [ isingomegafig ] ; notice that fig [ isingfig ] is quite similar to fig [ fermifig ] .
in all three regions of fig [ isingfig ] , at the largest frequencies , @xmath84 , @xmath116 displays the critical correlations of the @xmath3 point as depicted in fig [ isingomegafig ] ( we have returned now to the language of the quantum model @xmath116 ) . in regions
a and b there is a crossover from these critical fluctuations , at an energy scale @xmath14 , to the behavior of the ordered ( @xmath154 ) or paramagnetic ( @xmath155 ) ground state of @xmath116 ( see fig [ isingomegafig ] ) .
both ground states have a gap , and thermal fluctuations will lead to dilute gas of quasiparticle excitations .
we expect that an effective classical model ( like glauber @xcite dynamics , which is similar in spirit to the langevin models of hohenberg and halperin @xcite ) will provide a suitable description of these thermal fluctuations . in @xmath64 , on the ordered side ( @xmath156 ) , these quasiparticle excitations are the ` kink ' and ` anti - kink ' solitons ; even an infinitesimal concentration of these is sufficient to destroy long - range order at any finite temperature .
however , for @xmath157 , long - range order is not immediately destroyed : as a result there is finite temperature phase transition within region a where the magnetic moment disappears ; this transition will be in the universality class of the @xmath34-dimensional classical ising model @xcite . in the quantum - critical region c ,
the critical fluctuations are quenched by thermal effects at the energy scale @xmath6 ( see fig [ isingomegafig ] ; an early analysis of finite @xmath6 crossovers in the transverse - field ising model by suzuki @xcite failed to identify region c ) .
this quenching is completely universal and will be described explicitly below .
the system has had no chance to display any characteristic of either non - critical ground state at any frequency scale . at low frequencies ,
the system realizes a new quantum relaxational regime .
it is this regime which is really characteristic of the region c , and not the high frequency critical behavior which is present in all three regions .
it is in this sense that the name `` quantum - critical '' of region c is a misnomer . as in the case of the dilute fermi and bose gases considered earlier ,
the dynamic susceptibility @xmath158 will satisfy a universal scaling form over the regions of fig [ isingfig ] : @xmath159 in the following we will determine the leading term in @xmath160 in the quantum - critical region c _
i.e. _ we will present an exact expression for @xmath161 . in the classical model @xmath127 at @xmath126 and @xmath141 , we know from ( [ chiteq0 ] ) that the green s function @xmath162 ; we are using the label @xmath8 for spatial direction corresponding to imaginary time .
we can now use a remarkable result of cardy @xcite , which relies on the conformal invariance of this critical theory , to obtain an exact result for @xmath14 in a system with a finite @xmath147 : @xmath163 this result has been asserted earlier by an inspection of the partial differential equation satisfied by @xmath14 @xcite .
although results like ( [ gxt ] ) have been known to conformal field theorists for some time , they usually interchange the roles of @xmath7 and @xmath8 _ i.e. _ they consider systems of finite spatial length @xmath164 , and infinite temporal length , so that the system is in its ground state . for us ,
the spatial extent is infinite , and there is a finite length @xmath147 along the @xmath8 direction , with the correlator ( [ gxt ] ) periodic in @xmath8 with period @xmath147 ( such a perspective has also been discussed by shankar @xcite and by korepin _ et .
@xcite ) . note that it does nt really make sense to talk about the long imaginary time limit , @xmath165 .
however , after analytic continuation to real time , the long time limit of the quantum problem , @xmath166 or @xmath167 , is eminently sensible , and is precisely the new quantum relaxational regime that we wish to access .
the analytic continuation is a little more convenient in fourier space : we fourier transform ( [ gxt ] ) to obtain @xmath168 at the matsubara frequencies @xmath169 and then analytically continue to real frequencies ( there are some interesting subtleties in the fourier transform to @xmath168 and its analytic structure in the complex @xmath84 plane , which are discussed elsewhere @xcite ) .
this gives us the universal function @xmath160 in the quantum - critical region : @xmath170 we show a plot of @xmath171 in fig [ chifig ] .
this result is the finite @xmath6 version of fig [ chiteq0fig ] .
notice that the sharp features of fig [ chiteq0fig ] have been smoothed out on the scale @xmath6 , and there is non - zero absorption at all frequencies .
we can also observe the crossover as a function of frequency claimed earlier in fig [ isingomegafig ] .
notice that for @xmath172 there is a well - defined peak in @xmath171 ( fig [ chifig ] ) rather like the @xmath63 critical behavior of fig [ chiteq0fig ] .
however , for @xmath173 we cross - over to the quantum relaxational regime and the spectral density @xmath171 is similar to a lorentzian around @xmath174 . this relaxational behavior can be characterized by a relaxation rate @xmath175 defined as @xcite @xmath176 ( this is motivated by the phenomenological relaxational form @xmath177 ) . from ( [ chiscale ] ) and ( [ phichi ] )
we determine : @xmath178 where we have returned to physical units .
the ease with which this result was obtained belies ( i claim ) its remarkable nature .
notice that we are working in a closed hamiltonian system , evolving unitarily in time with the operator @xmath179 , from an initial density matrix given by the gibbs ensemble at a temperature @xmath6 .
yet , we have obtained relaxational behavior at low frequencies , and determined an exact value for a dissipation constant .
such behavior is more typically obtained in phenomenological models which couple the system to an external heat bath and postulate an equation of motion of the langevin type .
notice also that @xmath180 in ( [ hamising ] ) is known to be integrable with an infinite number of conservation laws @xcite . however , the conservation laws are associated with a mapping to a free fermion model and are highly non - local in our @xmath119 degrees of freedom ; they play essentially no role in our considerations , and do not preclude relaxational behavior in the @xmath119 variables . for completeness we also present results on a related observable which shows the crossover from critical to quantum relaxation behavior .
we consider the local susceptiblity @xmath181 , and obtain the finite @xmath6 form of ( [ chilteq0 ] ) by integrating ( [ phichi ] ) over momenta : @xmath182 a plot of the scaling function @xmath183 is shown in fig [ chilocfig ] .
the function has the asymptotic limits : @xmath184 the small @xmath185 behavior is relaxational as @xmath183 is linear in frequency , and the critical behavior at large frequencies agrees with ( [ chilteq0 ] ) .
the study of this model is of direct experimental interest , as it is believed @xcite to be a reasonable model of the spin fluctuations in antiferromagnetic compounds like @xmath186 and its lightly doped variants .
the insight gained from the simple models studied in the previous sections will now be of great use , and we will rapidly be able to present a scaling analysis of its quantum phase transition .
the hamiltonian of the quantum rotor model is @xmath187 where @xmath188 is an overall energy scale , @xmath189 is a dimensionless coupling constant , and @xmath33,@xmath190 are the sites of a two dimensional lattice ( @xmath35 denotes nearest neighbors ) .
notice the similarity between the forms of @xmath191 and @xmath116 in ( [ hamising ] ) : it will turn out that the corresponding terms play a similar role . on each site @xmath33 of the lattice we have the 3-component vector operators @xmath192 , @xmath193 ( dropping the site index ) , which obey the commutation relations : @xmath194 = 0~~,~~[l_a , n_b ] = i \epsilon_{abc } n_c~~,~~[l_a , l_b ] = i \epsilon_{abc } l_c .\ ] ] the vector @xmath193 is of unit length @xmath195 , and its orientation identifies direction of the local magnetic order ; the quantum rotor model is usually considered as an effective model for an underlying system of heisenberg spins in this case the magnetic order can be any ordering which is specified by a single vector and has no spatially averaged magnetic moment .
the simplest example of this is the two sublattice nel ordering , and we will therefore refer to @xmath196 as the nel order parameter .
the @xmath192 operator measures the angular momentum , and as all phases have no net magnetic moment , we will always have @xmath197 .
for further insight into the meaning of @xmath191 , consider the eigenstates of a single site hamiltonian @xmath198 .
this describes a particle moving on a unit sphere with angular co - ordinate @xmath193 and kinetic energy @xmath199 .
its eigenenergies are @xmath200 with degeneracy @xmath201 where @xmath202 .
the ground state is a non - degenerate singlet ( @xmath203 ) and has maximum uncertainty in the orientation of @xmath193 . for large @xmath1
, the ground state of @xmath191 can be approximated by the tensor product of @xmath204 states on each site .
this state is clearly a quantum paramagnet and has a gap , @xmath205 , to all excitations .
notice the similarity between this state and the large @xmath1 quantum paramagnet of the ising model @xmath116 ; in both cases the order parameters @xmath119 , @xmath193 are in a state of maximum uncertainty .
the small @xmath1 limit of @xmath191 is also similar to the small @xmath1 limit of @xmath116 : now the exchange interactions between the sites prefer a state in which @xmath193 has the same definite orientation on each site .
therefore , we expect long - range nel order in the small @xmath1 ground state . these two limiting states will be separated by a quantum phase transition at @xmath3 , which is , of course , the main subject of interest in this section . before discussing the critical properties
, we pause to remark on the relationship between the rotor model and heisenberg antiferromagnets . consider a pair of antiferromagnetically coupled spin-@xmath206 heisenberg spins : the eigenstates of this pair will have energies @xmath207 for @xmath208 .
notice the similarity between these states and those of a single quantum rotor ; the only difference is that there is no upper limit on the maximum value of @xmath209 in the rotor case .
however , it is reasonable to expect that these extra high energy states will not modify the low energy properties of lattice models .
therefore , there is little reason to doubt that the critical properties of heisenberg antiferromagnets with a natural pairing of spins ( or more generally , a natural clustering into an even number of spins ) will be same as those of the rotor model .
antiferromagnets with no such pairing contain net berry phase terms in their imaginary time path integral , beyond those present for the rotor model .
however these berry phases cancel between the sites , except for `` hedgehog''-like spacetime singularities @xcite , and it has been argued @xcite that these remnant berry phases have no effect of the leading critical singularities .
the reader is referred to the original papers for further discussion on these subtle issues @xcite : we will restrict our discussion here to the much simpler rotor model @xmath191 .
the critical properties and phase diagram of the @xmath30 rotor model @xmath191 turn out to be remarkably similar to those of the transverse field ising model @xmath116 of ( [ hamising ] ) in @xmath64 . like the @xmath64 ising model
, the @xmath30 rotor model has no phase transition at any finite @xmath6 : so the phase diagram of fig [ isingfig ] applies to @xmath210 , with no phase boundary in region a ( the phase diagram for @xmath191 in @xmath30 was obtained first by chakravarty _ et .
@xcite ) . in both systems ,
the @xmath63 , @xmath3 critical point has @xmath128 . in the case of @xmath116
this critical point was described by a cqft which upon analytic continuation to imaginary time was the field theory of the two - dimensional classical ising model .
the analogous mapping for @xmath191 yields the cqft associated with the field theory for the three - dimensional , classical , heisenberg ferromagnet .
the universal scaling functions describing the crossovers in fig [ isingfig ] and [ isingomegafig ] have an identical form in both theories , although , because the critical field theories are different , the critical exponents , universal amplitude ratios , and scaling functions will have different numerical values .
the explicit results presented in sec [ isingsec ] for quantum - critical region c of the ising model , all apply , unchanged in form , to the region c of the @xmath30 rotor model : the qualitative features of the spectral functions in figs [ chiteq0fig ] , [ chifig ] , and [ chilocfig ] remain the same , and the relaxation rate @xmath175 ( defined in ( [ gammadef ] ) ) satisfies ( [ gammares ] ) but with a different universal numerical prefactor .
however , unlike the @xmath64 ising model , we can not now get exact numerical results for the scaling functions of @xmath191 : this is because the three - dimensional classical hiesenberg ferromagnet is not exactly solvable ( unlike the two - dimensional classical ising model ) .
instead we have to be satisfied by approximate methods ; reasonably accurate numerical estimates can be obtained in the @xmath211 expansion which has been discussed at length by chubukov , sachdev and ye @xcite .
there is a small , but significant , difference between the ising model in @xmath64 and the rotor model in @xmath30 which can not go unmentioned .
this difference applies mainly to `` ordered '' regime in region a ( see figs [ isingfig ] and [ isingomegafig ] ) .
the @xmath63 ground state of @xmath116 for @xmath154 has a gap , associate with the finite energy cost of creating a kink or anti - kink soliton .
in contrast , the @xmath156 ground state of @xmath191 has gapless spin - wave excitations because of the broken continuous @xmath29 symmetry of the ordered state .
so we can no longer use the gap , @xmath205 , as the energy scale , @xmath212 for measuring deviations from @xmath3 for @xmath213 .
a convenient substitute turns out to be the spin stiffness @xmath214 which has the physical dimensions of energy in @xmath30 , and which also vanishes as @xmath215 . at finite @xmath6 in region a , it is known @xcite that the spin correlation length @xmath216 as @xmath217 .
it is interesting to note that the behavior of @xmath180 in @xmath64 is very similar : in this case the correlation length is determined by the mean spacing between kinks , and therefore behaves as @xmath218 for low @xmath6 in region a. ( related to the exponential divergence of the correlation length , there is a further sub - division of the `` ordered '' regime of region a ( figs [ isingfig ] and [ isingomegafig ] ) at the energy scale @xmath219 ; this complication occurs for both the @xmath30 rotor and the @xmath64 ising models , and has been discussed elsewhere @xcite . ) associated with the continuous @xmath29 symmetry of @xmath191 , there is an important observable whose properties can not be deduced by an analogy with the the ising model .
this is the uniform susceptibility @xmath220 , the response to a field , @xmath221 which couples to the global conserved charge associated with the continuous symmetry : @xmath222 the scaling dimension of @xmath220 can be determined exactly using symmetry arguments and the assumption of hyperscaling @xcite : this yields the scaling form @xmath223 here @xmath135 , is the same velocity that appears in a @xmath63 correlator like ( [ chiteq0 ] ) , and @xmath224 is a fully universal function .
this function was computed exactly @xcite in a @xmath225 rotor model , along with @xmath211 corrections in some limits ; it has the limiting behavior @xmath226 + \ldots & r \rightarrow 0 \\
\displaystyle \frac{2r}{n } + \frac{n-2}{n } + \ldots & r \rightarrow \infty \end{array}\right . \label{chihres}\ ] ] the two terms in the second result ( @xmath227 ) are expected to be exact to all orders in @xmath211 ; the same two terms were also obtained by hasenfratz and niedermayer @xcite .
notice from ( [ chihscale ] ) and ( [ chihres ] ) that @xmath220 has a linear dependence on @xmath6 both for @xmath228 and @xmath229 ; the slopes however differ by a factor of about 3 ( for @xmath230 ) and this will be important for experimental comparisons @xcite .
the most straightforward comparison is with the double - layer , spin-1/2 heisenberg antiferromagnet @xcite .
this model consists of spin-1/2 heisenberg spins on two adjacent square lattices , with an intralayer antiferromagnetic exchange @xmath231 and an interlayer antiferromagnetic exchange @xmath232 .
the ratio @xmath233 acts much like the dimensionless coupling @xmath1 , with the large @xmath233 a gapped quantum paramagnet of singlet pairs of spins in opposite layers , and the small @xmath233 magnetically ordered .
extensive numerical simulations have been carried out on this model by sandvik and collaborators @xcite , and the critical point @xmath126 identified rather precisely .
it is then possible to study the quantum - critical region c quite carefully as it extends over the maximum @xmath6 range .
a number of universal amplitude ratios , including those associated with @xmath220 , and all results are now in good agreement with the @xmath211 expansion on the @xmath234 quantum rotor model . results from high temperature series expansions on the double - layer model also support this conclusion @xcite .
secondly , comparisons have been made with the single - layer , square lattice spin-1/2 heisenberg antiferromagnet , both via simulations and by experimental measurements on @xmath235 .
this model has long - range order at @xmath63 , so must map onto the non - linear sigma model with @xmath156 .
the low @xmath6 region a was studied in the paper of chakravarty , halperin and nelson @xcite , with good experimental agreement .
here we focus on the issue of whether this lattice model exhibits the cqft high @xmath6 behavior of region c , or it goes directly from region a to a non - universal , lattice dominated high @xmath6 region like d of fig [ isingfig ] .
it was first argued by chubukov and sachdev @xcite that this model does indeed possess a significant intermediate temperature regime of region c : this was based on comparisons with the @xmath6 dependence of @xmath220 , with the factor of 3 alluded to above playing an important role .
they also noted that it would be difficult to identify this region in the correlation length , an observation that was subsequently re - iterated by greven _
. a rather convincing demonstration of the presence of region c was given recently by elstner _
@xcite who examined a large number of observables in a high @xmath6 expansion , and found good consistency with the universal rotor model results . also significant in this context have been nuclear magnetic resonance experiments of imai _ et .
_ @xcite on @xmath186 .
they have measured the @xmath6 dependence of the @xmath236 and @xmath237 relaxation rates at intermediate temperatures .
their observations are in reasonable agreement with the predictions @xcite that can be obtained from the universal scaling result for @xmath238 in the quantum - critical region c.
this paper has presented a discussion of the vicinity of a second - order quantum phase transition in the context of a number of simple models . the overall picture that emerges
is summarized in fig [ generalfig ] , which shows a generic phase diagram in the plane of a coupling constant @xmath1 , and the temperature @xmath6 of a @xmath34-dimensional system ; this phase diagram is a generalization of a diagram obtained first by chakravarty _ et .
@xcite for the @xmath30 quantum rotor model .
the quantum phase transition occurs at the point @xmath3 , @xmath63 .
associated with this critical point , we can define a continuum quantum field theory ( cqft ) over spacetime .
in general , space ( @xmath7 ) and time ( @xmath8 ) do _ not _ play a similar role in the cqft and have different scaling dimensions : @xmath239 , @xmath240 under a spatial rescaling by @xmath241 , with @xmath10 the dynamic exponent .
further , even in imaginary time , the action for the cqft can be complex due to the presence of berry phases , and therefore corresponds to a statistical mechanics model with complex weights .
correlators of the cqft are the universal functions describing crossovers in the vicinity of @xmath3 , @xmath63 .
the cqft has no ultraviolet cutoff , but is characterized solely by two energy scales : the temperature @xmath6 and an energy scale @xmath242 characterizing the deviation of the ground state from the critical point ( here @xmath13 is the correlation length exponent ) .
the value of the ratio @xmath21 determines two distinct regions of the cqft shown in fig [ generalfig ] . in both regions
there is a high frequency regime ( @xmath243 ) which is dominated by excitations of the critical cqft of the @xmath3 , @xmath63 point .
the regions are distinguished only by their low frequency behavior , which we discuss in turn : ( _ i _ ) the low @xmath6 region ( @xmath244 , shown shaded in fig [ generalfig ] ) is a region of `` conventional '' physics for frequency scales @xmath245 .
this frequency regime can be understood by beginning with the non - critical ground state and examining the particle - like excitations above it . a simple classical model ( a boltzmann equation for a gas of quasiparticle excitations , or a langevin model of the types discussed by hohenberg and halperin @xcite ) is usually adequate for describing the long - distance , long - time dynamics of these excitations .
the shaded region can also contain thermally driven , phase transitions ; these transitions will be described by a classical field theory .
( _ ii _ ) the high @xmath6 region of the cqft has a novel quantum relaxational regime .
this regime is not described by an effective classical model , and displays intrinsic quantum - mechanical effects at the longest time and distance scales .
a scaling analysis for this regime was reviewed in this paper . in some cases , as in the model of section [ isingsec ] , it is possible to obtain an exact value for the relaxation constant .
* quantum transitions between fermi liquids and states with various types of spin or charge density wave orderings were discussed in important early work by hertz @xcite .
he focussed on the immediate vicinity of the finite termperature transition , like that within region a in fig [ isingfig ] .
in particular , hertz missed the existence of the `` quantum - critical '' regime ( as was pointed out recently by millis @xcite this oversight is similar to suzuki s
@xcite for the ising model ) , which has been the main focus of this paper .
more detailed studies of quantum transitions involving fermi liquids have appeared recently @xcite . a related , but different ,
perspective is provided by studies of critical phenomena in rotor models with doped electrons @xcite . *
related ideas on scaling in the quantum critical region have been presented by tsvelik and collaborators @xcite . *
a great deal of work has been done recently on `` quantum impurity '' models @xcite like the multi - channel kondo effect .
these models also display quantum phase transitions , with crossovers bearing some similarity to those discussed here .
however the transitions do not modify the bulk properties , and are related instead to boundary critical phenomena . * as we indicated briefly in the discussion on the bose gas in section [ sec : bose ] , dangerously irrelevant operators sometimes need to be considered , as they do in the dilute bose gas for @xmath246 .
in fact such effects arise somewhat more frequently than they do in classical critical phenomena , as the upper critical dimension of the quantum transition is often quite low .
an early analysis of the dilute bose gas in @xmath247 by weichmann _ et .
@xcite was dominated by such effects , although they did not present their results in the general context of quantum phase transitions .
such a perspective can be found in more recent work @xcite .
* an important subject , on which much is not understood , is the effect of quenched randomness on quantum phase transitions .
an early analysis for the random quantum rotor model was given by boyanovsky and cardy @xcite .
a recent exact solution by fisher @xcite of the random transverse - field ising model in @xmath64 represents significant progress .
the nature of spin glass ordering at @xmath63 and its destruction by quantum fluctuations has also been studied recently @xcite . *
a number of experiments @xcite have reported scaling of the type in eqn ( [ phildef ] ) in the local dynamic susceptibility .
however the universality classes controlling these systems are not understood and quenched randomness appears to play a significant role . nevertheless , it is interesting to note the qualitative similarity between the experimental measurements in fig . 4 of aronson _ et .
al._@xcite and our result for the local susceptibility in fig [ chilocfig ] .
the latter measures spin correlations on a single ising spin induced via its coupling to its environment of other ising spins .
however , there is a fundamental equivalence between the spin being measured and its environment ; this is an important difference between our bulk approach and alternative descriptions of the experiments using `` quantum impurity '' models @xcite which clearly distinguish between the impurity and environment degrees of freedom . *
a recent study @xcite has examined the cqft of a quantum ferromagnet .
this system does not display a quantum phase transition of the type discussed here .
nevertheless , its phase diagram has regions similar to those in fig [ generalfig ] and aspects of its scaling properties are related to those of the dilute bose gas of section [ sec : bose ] .
i thank n. read , r. shankar , a. sokol , t. senthil , j. ye , and especially a.v .
chubukov for collaborations on the topics reviewed in this paper .
i am grateful to a.v .
chubukov and t. senthil for valuable comments on the manuscript and r.e .
shrock for helpful discussions .
this research was supported by national science foundation grant dmr-9224290 .
e. lieb , t. schultz , and d. mattis , ann . of phys .
* 16 * , 406 ( 1961 ) ; th .
niemeijer , physica * 36 * , 377 ( 1967 ) , * 39 * , 313 ( 1968 ) ; p. pfeuty , ann . of phys .
* 57 * , 79 ( 1970 ) ; e. barouch and b.m .
mccoy , phys .
b * 3 * , 786 ( 1971 ) ; b.m .
mccoy , j.h.h .
perk , and r.e .
shrock , nucl .
b*220 * , 35 ( 1983 ) . | we study the finite temperature crossovers in the vicinity of a zero temperature quantum phase transition .
the universal crossover functions are observables of a continuum quantum field theory .
particular attention is focussed on the high temperature limit of the continuum field theory , the so - called `` quantum - critical '' region .
basic features of crossovers are illustrated by a simple solvable model of dilute spinless fermions , and a partially solvable model of dilute bosons .
the low frequency relaxational behavior of the quantum - critical region is displayed in the solution of the transverse - field ising model .
the insights from these simple models lead to a fairly complete understanding of the system of primary interest : the two - dimensional quantum rotor model , whose phase transition is expected to be in the same universality class as those in antiferromagnetic heisenberg spin models .
recent work on the experimental implications of these results for the cuprate compounds is briefly reviewed .
proceedings of the + 19th iupap international conference on statistical physics + xiamen , china , july 31 - august 4 1995 + world scientific , to be published , edited by b .-
l .
hao . |
content on the internet is becoming increasingly multilingual .
a prime example is wikipedia . in 2001 , the majority of pages were written in english , while in 2015 , the percentage of english articles has dropped to 14% . at the same time , online news has begun to dominate reporting of current events .
however , machine translation remains relatively rudimentary .
it allows people to understand simple phrases on web pages , but remains inadequate for more advanced understanding of text . in this paper
we consider the intersection of these developments : how to track events which are reported about in multiple languages .
the term event is vague and ambiguous , but for the practical purposes , we define it as `` any significant happening that is being reported about in the media .
'' examples of events would include shooting down of the malaysia airlines plane over ukraine on july 18th , 2014 and hsbc s admittance of aiding their clients in tax evasion on february 9th , 2015 ( figure [ fig : event2 ] ) .
events such as these are covered by many articles and the question is how to find all the articles in different languages that are describing a single event .
generally , events are more specific than general themes as the time component plays an important role for example , the two wars in iraq would be considered as separate events .
events are represented by collections of articles about an event , in this case the malaysian airliner which was shot down over ukraine .
the results shown in the figure can be obtained using the query http://eventregistry.org/event/997350#?lang=eng&tab=articles .
the content presented is part of the event registry system , developed by the authors.,scaledwidth=100.0% ] as input , we consider a stream of articles in different languages and a list of events . our goal is to assign articles to their corresponding events .
a priori , we do not know the coverage of the articles , that is , not all the events may be covered and we do not know that all the articles necessarily fit into one of the events .
the task is divided into two parts : detecting events within each language and then linking events across languages . in this paper
we address the second step .
we consider a high volume of articles in different languages . by using a language detector ,
the stream is split into separate monolingual streams . within each monolingual stream ,
an online clustering approach is employed , where tracked clusters correspond to our definition of events - this is based on the event registry system @xcite .
our main goal in this paper is to connect such clusters ( representations of events ) across languages , that is , to detect that a set of articles in language @xmath0 reports on the same event as a set of articles in language @xmath1 .
our approach to link clusters across languages combines two ingredients : a cross - lingual document similarity measure , which can be interpreted as a language independent topic model , and semantic annotation of documents , which enables an alternative way to comparing documents . since this work represents a complicated pipeline ,
we concentrate on these two specific elements .
overall , the approach should be considered from a systems perspective ( considering the system as a whole ) rather than considering these problems in isolation .
the first ingredient of our approach to link clusters across languages represents a continuation of previous work @xcite where we explored representations of documents which were valid over multiple languages .
the representations could be interpreted as multilingual topics , which were then used as proxies to compute cross - lingual similarities between documents . to learn the representations , we use wikipedia as a training corpus .
significantly , we do not only consider the major or _ hub _ languages such as english , german , french , etc . which have significant overlap in article coverage , but also smaller languages ( in terms of number of wikipedia articles ) such as slovenian and hindi , which may have a negligible overlap in article coverage .
we can then define a similarity between any two articles regardless of language , which allows us to cluster the articles according to topic .
the underlying assumption is that articles describing the same event are similar and will therefore be put into the same cluster .
based on the similarity function , we propose a novel algorithm for linking events / clusters across languages .
the approach is based on learning a classification model from labelled data based on several sets of features .
in addition to these features , cross - lingual similarity is also used to quickly identify a small list of potential linking candidates for each cluster .
this greatly increases the scalability of the system .
the paper is organized as follows : we first provide an overview of the system as a whole in section [ sec : pipeline ] , which includes a subsection that summarizes the main system requirements .
we then present related work in section [ sec : related ] .
the related work covers work on cross - lingual document similarity as well as work on cross - lingual cluster linking . in section [ sec : crosslingual ] , we introduce the problem of cross - lingual document similarity computation and describe several approaches to the problem , most notably a new approach based on hub languages . in section [ sec : linking ] , we introduce the central problem of cross - lingual linking of clusters of news articles and our approach that combines the cross - lingual similarity functions with knowledge extraction based techniques .
finally , we present and interpret the experimental results in section [ sec : evaluation ] and discuss conclusions and point out several promising future directions .
we base our techniques of cross - lingual event linking on an online system for detection of world events , called event registry @xcite .
event registry is a repository of events , where events are automatically identified by analyzing news articles that are collected from numerous news outlets all over the world .
the important components in the pipeline of the event registry are shown in figure [ fig : erpipeline ] .
we will now briefly describe the main components . the event registry pipeline .
after new articles are collected , they are first analyzed individually ( article - level processing ) . in the next step ,
groups of articles about the same event are identified and relevant information about the event is extracted ( event construction phase ) .
although the pipeline contains several components , we focus only on the two highlighted in the image . ]
the collection of the news articles is performed using the newsfeed service @xcite .
the service monitors rss feeds of around 100,000 mainstream news outlets available globally . whenever a new article is detected in the rss feed
, the service downloads all available information about the article and sends the article through the pipeline .
newsfeed downloads daily on average around 200,000 news articles in various languages , where english , spanish and german are the most common .
collected articles are first semantically annotated by identifying mentions of relevant concepts either entities or important keywords .
the disambiguation and entity linking of the concepts is done using wikipedia as the main knowledge base .
the algorithm for semantic annotation uses machine learning to detect significant terms within unstructured text and link them to the appropriate wikipedia articles .
the approach models link probability and combines prior word sense distributions with context based sense distributions .
the details are reported in @xcite and @xcite . as a part of the semantic annotation we also analyze the dateline of the article to identify the location of the described event as well as to extract dates mentioned in the article using a set of regular expressions .
since articles are frequently revised we also detect if the collected article is just a revision of a previous one so that we can use this information in next phases of the pipeline .
the last important processing step on the document level is to efficiently identify which articles in other available languages are most similar to this article .
the methodology for this task is one of the main contributions of this paper and is explained in details in section [ sec : crosslingual ] . as the next step ,
an online clustering algorithm is applied to the articles in order to identify groups of articles that are discussing the same event . for each new article
, the clustering algorithm determines if the article should be assigned to some existing cluster or into a new cluster .
the underlying assumption is that articles that are describing the same event are similar enough and will therefore be put into the same cluster . for clustering , each new article is first tokenized , stop words are removed and the remaining words are stemmed .
the remaining tokens are represented in a vector - space model and normalized using tf - idf ( see section [ sec : tfidf ] for the definition ) .
cosine similarity is used to find the most similar existing cluster , by comparing the document s vector to the centroid vector of each cluster .
a user - defined threshold is used to determine if the article is not similar enough to any existing clusters ( 0.4 was used in our experiments ) .
if the highest similarity is above the threshold , the article is assigned to the corresponding cluster , otherwise a new cluster is created , initially containing only the single article . whenever an article is assigned to a cluster ,
the cluster s centroid vector is also updated . since articles about an event
are commonly written only for a short period of time , we remove clusters once the oldest article in the cluster becomes more than 4 days old .
this housekeeping mechanism prevents the clustering from becoming slow and also ensures that articles are not assigned to obsolete clusters .
details of the clustering approach are described in @xcite .
once the number of articles in a cluster reaches a threshold ( which is a language dependent parameter ) , we assume that the articles in the cluster are describing an event . at that point , a new event with a unique i d is created in event registry , and the cluster with the articles is assigned to it . by analyzing the articles , we extract the main information about the event , such as the event location , date , most relevant entities and keywords , etc .
since articles in a cluster are in a single language , we also want to identify any other existing clusters that report about the same event in other languages and join these clusters into the same event .
this task is performed using a classification approach which is the second major contribution of this paper .
it is described in detail in section [ sec : linking ] . when a cluster is identified and information about the event is extracted , all available data is stored in a custom - built database system .
the data is then accessible through the api or a web interface at http://eventregistry.org/ , which provide numerous search and visualization options .
our goal is to build a system that monitors global media and analyzes how events are being reported on .
our approach consists of two steps : tracking events separately in each language ( based on language detection and an online clustering approach ) and then connecting them .
the pipeline must be able to process millions of articles per day and perform billions of similarity computations each day .
both steps rely heavily on similarity computation , which implies that this must be highly scalable . therefore , we focus on implementations that run on a single shared memory machine , as opposed to clusters of machines .
this simplifies implementation and system maintenance .
to summarize , the following properties are desirable : * * training * - the training ( building cross - lingual models ) should scale to many languages and should be robust to the quality of training resources .
the system should be able to take advantage of comparable corpora ( as opposed to parallel translation - based corpora ) , with missing data .
* * operation efficiency * - the similarity computation should be fast - the system must be able to handle billions of similarity computations per day
. computing the similarity between a new document and a set of known documents should be efficient ( the main application is linking documents between two monolingual streams ) . * * operation cost * - the system should run on a strong shared machine server and not rely on paid services . * * implementation * - the system is straightforward to implement , with few parameters to tune .
we believe that a cross - lingual similarity component that meets such requirements is very desirable in a commercial setting , where several different costs have to be taken into consideration .
in this section , we describe previous work described in the literature .
since there are two distinctive tasks that we tackle in this paper ( computing cross - lingual document similarity and cross - lingual cluster linking ) , we have separated the related work into two corresponding parts .
there are four main families of approaches to cross - lingual similarity .
* translation and dictionary based*. the most obvious way to compare documents written in different languages is to use machine translation and perform monolingual similarity , see @xcite for several variations of translation based approaches .
one can use free tools such as moses @xcite or translation services , such as google translate ( https://translate.google.com/ ) .
there are two issues with such approaches : they solve a harder problem than needs to be solved and they are less robust to training resource quality - large sets of translated sentences are typically needed .
training moses for languages with scarce linguistic resources is thus problematic .
the issue with using online services such as google translate is that the apis are limited and not free .
the operation efficiency and cost requirements make translation - based approaches less suited for our system .
closely related are works cross - lingual vector space model ( cl - vsm ) @xcite and the approach presented in @xcite which both compare documents by using dictionaries , which in both cases are eurovoc dictionaries @xcite .
the generality of such approaches is limited by the quality of available linguistic resources , which may be scarce or non - existent for certain language pairs .
* probabilistic topic models*. there exist many variants to modelling documents in a language independent way by using probabilistic graphical models .
the models include : joint probabilistic latent semantic analysis ( jplsa ) @xcite , coupled probabilistic lsa ( cplsa ) @xcite , probabilistic cross - lingual lsa ( pcllsa ) @xcite and polylingual topic models ( pltm ) @xcite which is a bayesian version of pcllsa . the methods ( except for cplsa ) describe the multilingual document collections as samples from generative probabilistic models , with variations on the assumptions on the model structure .
the topics represent latent variables that are used to generate observed variables ( words ) , a process specific to each language .
the parameter estimation is posed as an inference problem which is typically intractable and one usually solves it using approximate techniques .
most variants of solutions are based on gibbs sampling or variational inference , which are nontrivial to implement and may require an experienced practitioner to be applied .
furthermore , representing a new document as a mixture of topics is another potentially hard inference problem which must be solved . *
matrix factorization*. several matrix factorization based approaches exist in the literature .
the models include : non - negative matrix factorization based @xcite , cross - lingual latent semantic indexing cl - lsi @xcite , canonical correlation analysis ( cca ) @xcite , oriented principal component analysis ( opca ) @xcite . the quadratic time and space dependency of the opca method makes it impractical for large scale purposes .
in addition , opca forces the vocabulary sizes for all languages to be the same , which is less intuitive . for our setting , the method in @xcite
has a prohibitively high computational cost when building models ( it uses dense matrices whose dimensions are a product of the training set size and the vocabulary size ) .
our proposed approach combines cca and cl - lsi .
another closely related method is cross - lingual explicit semantic analysis ( cl - esa ) @xcite , which uses wikipedia ( as do we in the current work ) to compare documents .
it can be interpreted as using the sample covariance matrix between features of two languages to define the dot product which is used to compute similarities .
the authors of cl - esa compare it to cl - lsi and find that cl - lsi can outperform cl - esa in an information retrieval , but is costlier to optimize over a large corpus ( cl - esa requires no training ) .
we find that the scalability argument does not apply in our case : based on advances in numerical linear algebra we can solve large cl - lsi problems ( millions of documents as opposed to the 10,000 document limit reported in @xcite ) . in addition , cl - esa is less suited for computing similarities between two large monolingual streams .
for example , each day we have to compute similarities between 500,000 english and 500,000 german news articles . comparing each german news article with 500,000 english news articles
is either prohibitively slow ( involves projecting all english articles on wikipedia ) or consumes too much memory ( involves storing the projected english articles , which for a wikipedia of size 1,000,000 is a 500,000 by 1000,0000 non - sparse matrix ) .
* monolingual*. finally , related work includes monolingual approaches that treat document written in different languages in a monolingual fashion .
the intuition is that named entities ( for example , `` obama '' ) and cognate words ( for example , `` tsunami '' ) are written in the same or similar fashion in many languages .
for example , the cross - language character n - gram model ( cl - cng ) @xcite represents documents as bags of character @xmath2-grams .
another approach is to use language dependent keyword lists based on cognate words @xcite .
these approaches may be suitable for comparing documents written in languages that share a writing system , which does not apply to the case of global news tracking . based on our requirements in section [ sec :
sysreq ] , we chose to focus on methods based on vector space models and linear embeddings .
we propose a method that is more efficient than popular alternatives ( a clustering - based approach and latent semantic indexing ) , but is still simple to optimize and use .
although there are a number of services that aggregate news by identifying clusters of similar articles , there are almost no services that provide linking of clusters over different languages .
google news as well as yahoo ! news are able to identify clusters of articles about the same event , but offer no linking of clusters across languages .
the only service that we found , which provides cross - lingual cluster linking , is the european media monitor ( emm ) @xcite .
emm clusters articles in 60 languages and then tries to determine which clusters of articles in different languages describe the same event .
to achieve cluster linking , emm uses three different language independent vector representations for each cluster .
the first vector contains the weighted list of references to countries mentioned in the articles , while the second vector contains the weighted list of mentioned people and organizations .
the last vector contains the weighted list of eurovoc subject domain descriptors .
these descriptors are topics , such as _ air transport _ , _ ec agreement _ , _ competition _ and _ pollution control _ into which articles are automatically categorized @xcite .
similarity between clusters is then computed using a linear combination of the cosine similarities computed on the three vectors .
if the similarity is above the threshold , the clusters are linked .
compared to emm , our approach uses document similarities to obtain a small set of potentially equivalent clusters .
additionally , we do not decide if two clusters are equivalent based on a hand - set threshold on a similarity value
instead we use a classification model that uses a larger set of features related to the tested pair of clusters .
a system , which is significantly different but worth mentioning , is the gdelt project @xcite . in gdelt
, events are also extracted from articles , but in their case , an event is specified in a form of a triple containing two actors and a relation .
the project contains an extensive vocabulary of possible relations , mostly related to political events . in order to identify events
, gdelt collects articles in more than 65 languages and uses machine translation to translate them to english .
all information extraction is then done on the translated article .
document similarity is an important component in techniques from text mining and natural language processing .
many techniques use the similarity as a black box , e.g. , a kernel in support vector machines .
comparison of documents ( or other types of text snippets ) in a monolingual setting is a well - studied problem in the field of information retrieval @xcite .
we first formally introduce the problem followed by a description of our approach .
we will first describe how documents are represented as vectors and how to compare documents in a mono - lingual setting .
we then define a way to measure cross - lingual similarity which is natural for the models we consider .
* document representation . *
the standard vector space model @xcite represents documents as vectors , where each term corresponds to a word or a phrase in a fixed vocabulary .
formally , document @xmath3 is represented by a vector @xmath4 , where @xmath2 corresponds to the size of the vocabulary , and vector elements
@xmath5 correspond to the number of times term @xmath6 occurred in the document , also called _ term frequency _ or @xmath7 .
we also used a term re - weighting scheme that adjusts for the fact that some words occur more frequently in general .
a term weight should correspond to the importance of the term for the given corpus .
the common weighting scheme is called _ term frequency inverse document frequency _ ( @xmath8 ) weighting .
an _ inverse document frequency _ ( @xmath9 ) weight for the dictionary term @xmath6 is defined as @xmath10 , where @xmath11 is the number of documents in the corpus which contain term @xmath6 . when building cross - lingual models , the idf scores were computed with respect to the wikipedia corpus . in the other part of our system
, we computed tfidf vectors on streams of news articles in multiple languages . there
the idf scores for each language changed dynamically - for each new document we computed the idf of all news articles within a 10 day window .
therefore we can define a document s @xmath8 as @xmath12 the @xmath8 weighted vector space model document representation corresponds to a map @xmath13 defined by : @xmath14 * mono - lingual similarity . *
a common way of computing similarity between documents is _ cosine similarity _ , @xmath15 where @xmath16 and @xmath17 are standard inner product and euclidean norm .
when dealing with two or more languages , one could ignore the language information and build a vector space using the union of tokens over the languages .
a cosine similarity function in such a space can be useful to some extent , for example `` internet '' or `` obama '' may appear both in spanish and english texts and the presence of such terms in both an english and a spanish document would contribute to their similarity .
in general however , large parts of vocabularies may not intersect .
this means that given a language pair , many words in both languages can not contribute to the similarity score .
such cases can make the similarity function very insensitive to the data .
* cross - lingual similarity .
* processing a multilingual dataset results in several vector spaces with varying dimensionality , one for each language .
the dimensionality of the vector space corresponding to the @xmath18-th language is denoted by @xmath19 and the vector space model mapping is denoted by @xmath20 .
the similarity between documents in language @xmath18 and language @xmath21 is defined as a bilinear operator represented as a matrix @xmath22 : @xmath23 where @xmath24 and @xmath25 are documents written in the @xmath18-th and @xmath21-th language respectively .
if the maximal singular value of @xmath26 is bounded by @xmath27 , then the similarity scores will lie on the interval @xmath28 $ ] .
we will provide an overview of the models in section [ sec : models ] and then introduce additional notation in [ sec : notation ] . starting with section [ sec : kmeans ] and ending with section [ sec : hublang ]
we will describe some approaches to compute @xmath26 given training data . in this section
, we will describe several approaches to the problem of computing the multilingual similarities introduced in section [ sec : tfidf ] .
we present four approaches : a simple approach based on @xmath6-means clustering in section [ sec : kmeans ] , a standard approach based on singular value decomposition in section [ sec : lsi ] , a related approach called canonical correlation analysis ( cca ) in section [ sec : cca ] and finally a new method , which is an extension of cca to more than two languages in section [ sec : hublang ] .
cca can be used to find correlated patterns for a pair of languages , whereas the extended method optimizes a sum of squared correlations ( sscor ) between several language pairs , which was introduced in @xcite .
the sscor problem is difficult to solve in our setting ( hundreds of thousands of features , hundreds of thousands of examples ) . to tackle this ,
we propose a method which consists of two ingredients .
the first one is based on an observation that certain datasets ( such as wikipedia ) are biased towards one language ( english for wikipedia ) , which can be exploited to reformulate a difficult optimization problem as an eigenvector problem .
the second ingredient is dimensionality reduction using cl - lsi , which makes the eigenvector problem computationally and numerically tractable .
we concentrate on approaches that are based on linear maps rather than alternatives , such as machine translation and probabilistic models , as discussed in the section on related work
. we will start by introducing some notation .
the cross - lingual similarity models presented in this paper are based on comparable corpora .
a _ comparable corpus _ is a collection of documents in multiple languages , with alignment between documents that are of the same topic , or even a rough translation of each other .
wikipedia is an example of a comparable corpus , where a specific entry can be described in multiple languages ( e.g. , berlin " is currently described in 222 languages ) .
news articles represent another example , where the same event can be described by newspapers in several languages . more formally , a _
multilingual document _
@xmath29 is a tuple of @xmath30 documents on the same topic ( comparable ) , where @xmath31 is the document written in language @xmath18 .
note that an individual document @xmath31 can be an empty document ( missing resource ) and each @xmath3 must contain at least * two nonempty documents*. this means that in our analysis we discard strictly monolingual documents for which no cross - lingual information is available .
a comparable corpus @xmath32 is a collection of @xmath33 multilingual documents . by using the vector space model
, we can represent @xmath34 as a set of @xmath30 matrices @xmath35 , where @xmath36 is the matrix corresponding to the language @xmath18 and @xmath19 is the vocabulary size of language @xmath18 .
furthermore , let @xmath37 denote the @xmath38-th column of matrix @xmath39 and the matrices respect the document alignment - the vector @xmath40 corresponds to the tfidf vector of the @xmath18-th component of multilingual document @xmath41 .
we use @xmath42 to denote the total row dimension of @xmath43 , i.e. , @xmath44 .
see figure [ fig : stacked_matrices ] for an illustration of the introduced notation .
multilingual corpora and their matrix representations using the vector space model.,width=340 ] we will now describe four models to cross - lingual similarity computation in the next sub - sections .
the @xmath6-means algorithm is perhaps the most well - known and widely - used clustering algorithm . here , we present its application to compute cross - lingual similarities .
the idea is based on concatenating the corpus matrices , running standard @xmath6-means clustering to obtain the matrix of centroids , reversing " the concatenation step to obtain a set of aligned bases , which are finally used to compute cross - lingual similarities .
see figure [ fig : kmeans ] for overview of the procedure .
the left side of figure [ fig : kmeans ] illustrates the decomposition and the right side summarizes the coordinate change .
@xmath6-means algorithm and coordinate change.,width=377 ] in order to apply the algorithm , we first merge all the term - document matrices into a single matrix @xmath43 by stacking the individual term - document matrices ( as seen in figure [ fig : stacked_matrices ] ) : @xmath45 such that the columns respect the alignment of the documents ( here matlab notation for concatenating matrices is used ) .
therefore , each document is represented by a long vector indexed by the terms in all languages .
we then run the @xmath6-means algorithm @xcite and obtain a centroid matrix @xmath46 , where the @xmath6 columns represent centroid vectors .
the centroid matrix can be split vertically into @xmath30 blocks : @xmath47^t,\ ] ] according to the number of dimensions of each language , i.e. , @xmath48 .
to reiterate , the matrices @xmath49 are computed using a multilingual corpus matrix @xmath43 ( based on wikipedia for example ) . to compute cross - lingual document similarities on new documents , note that each matrix @xmath49 represents a vector space basis and can be used to map points in @xmath50 into a @xmath6-dimensional space , where the new coordinates of a vector @xmath51 are expressed as : @xmath52 the resulting matrix for similarity computation between language @xmath18 and language @xmath21 is defined up to a scaling factor as : @xmath53 the matrix is a result of mapping documents in a language independent space using pseudo - inverses of the centroid matrices @xmath54 and then comparing them using the standard inner product , which results in the matrix @xmath55 . for the sake of presentation , we assumed that the centroid vectors are linearly independent .
( an independent subspace could be obtained using an additional gram - schmidt step @xcite on the matrix @xmath56 , if this was not the case . ) the second approach we consider is cross - lingual latent semantic indexing ( cl - lsi ) @xcite which is a variant of lsi @xcite for more than one language .
the approach is very similar to @xmath6-means , where we first concatenate the corpus matrices , compute a decomposition , which in case of cl - lsi is a truncated singular value decomposition ( svd ) , decouple the column space matrix and use the blocks to compute linear maps to a common vector space , where standard cosine similarity is used to compare documents .
the method is based on computing a truncated singular value decomposition of the concatenated corpus matrix @xmath57 .
see figure [ fig : lsi ] for the decomposition . representing documents in topic
coordinates is done in the same way as in the @xmath6-means case ( see figure [ fig : kmeans ] ) , we will describe how to compute the coordinate change functions .
lsi multilingual corpus matrix decomposition.,width=377 ] the cross - lingual similarity functions are based on a rank-@xmath6 truncated svd : @xmath58 where @xmath59 are basis vectors of interest and @xmath60 is a truncated diagonal matrix of singular eigenvalues .
an aligned basis is obtained by first splitting @xmath61 vertically according to the number of dimensions of each language : @xmath62^t$ ] .
then , the same as with @xmath6-means clustering , we compute the pseudoinverses @xmath63 .
the matrices @xmath64 are used to change the basis from the standard basis in @xmath50 to the basis spanned by the columns of @xmath65 .
[ [ implementation - note ] ] implementation note + + + + + + + + + + + + + + + + + + + since the matrix @xmath43 can be large we could use an iterative method like the lanczos algorithm with reorthogonalization @xcite to find the left singular vectors ( columns of @xmath61 ) corresponding to the largest singular values .
it turns out that the lanczos method converges slowly as the gap between the leading singular values is small .
moreover , the lanczos method is hard to parallelize .
instead , we use a randomized version of the svd @xcite that can be viewed as a block lanczos method .
that enables us to use parallelization and speeds up the computation considerably . to compute the matrices @xmath64 we used the qr algorithm @xcite to factorize @xmath65 as @xmath66 , where @xmath67 and @xmath68 is a triangular matrix
@xmath64 is then obtained by solving @xmath69 .
we now present a statistical technique to analyze data from two sources , an extension of which will be presented in the next section .
canonical correlation analysis ( cca ) @xcite is a dimensionality reduction technique similar to principal component analysis ( pca ) @xcite , with the additional assumption that the data consists of feature vectors that arose from two sources ( two views ) that share some information .
examples include : bilingual document collection @xcite and collection of images and captions @xcite . instead of looking for linear combinations of features that maximize the variance ( pca )
we look for a linear combination of feature vectors from the first view and a linear combination for the second view , that are maximally correlated . interpreting the columns of @xmath39 as observation vectors sampled from an underlying distribution @xmath70 ,
the idea is to find two weight vectors @xmath71 and @xmath72 so that the random variables @xmath73 and @xmath74 are maximally correlated ( @xmath75 and @xmath76 are used to map the random vectors to random variables , by computing weighted sums of vector components ) .
let @xmath77 denote the sample - based correlation coefficient between two vectors of observations @xmath78 and @xmath79 . by using the sample matrix notation @xmath39 and
@xmath80 ( assuming no data is missing to simplify the exposition ) , this problem can be formulated as the following optimization problem : @xmath81 where @xmath82 and @xmath83 are empirical estimates of variances of @xmath84 and @xmath85 respectively and @xmath86 is an estimate for the covariance matrix . assuming that the observation vectors are centered ( only for the purposes of presentation ) , the matrices are computed in the following way : @xmath87 , and similarly for @xmath82 and @xmath83 . the optimization problem can be reduced to an eigenvalue problem and includes inverting the variance matrices @xmath82 and @xmath83 .
if the matrices are not invertible , one can use a regularization technique by replacing @xmath82 with @xmath88 , where @xmath89 $ ] is the regularization coefficient and @xmath90 is the identity matrix .
( the same can be applied to @xmath83 . )
a single canonical variable is usually inadequate in representing the original random vector and typically one looks for @xmath6 projection pairs @xmath91 , so that @xmath92 and @xmath93 are highly correlated and @xmath92 is uncorrelated with @xmath94 for @xmath95 and analogously for @xmath96 vectors .
note that the method in its original form is only applicable to two languages where an aligned set of observations is available .
the next section will describe a scalable extension of cca to more than two languages . building cross - lingual similarity models based on comparable corpora is challenging for two main reasons .
the first problem is related to missing alignment data : when a number of languages is large , the dataset of documents that cover all languages is small ( or may even be empty ) .
even if only two languages are considered , the set of aligned documents can be small ( an extreme example is given by the piedmontese and hindi wikipedias where no inter - language links are available ) , in which case none of the methods presented so far are applicable .
the second challenge is scale - the data is high - dimensional ( many languages with hundreds of thousands of features per language ) and the number of multilingual documents may be large ( over one million in case of wikipedia ) .
the optimization problem posed by cca is not trivial to solve : the covariance matrices themselves are prohibitively large to fit in memory ( even storing a 100,000 by 100,000 element matrix requires 80 gb of memory ) and iterative matrix - multiplication based approaches to solving generalized eigenvalue problems are required ( the covariance matrices can be expressed as products of sparse matrices , which means we have fast matrix - vector multiplication ) .
we now describe an extension of cca to more than two languages , which can be trained on large comparable corpora and can handle missing data .
the extension we consider is based on a generalization of cca to more than two views , introduced in @xcite , namely the sum of squared correlations sscor , which we will state formally later in this section .
our approach exploits a certain characteristic of the data , namely the _ hub language _ characteristic ( see below ) in two ways : to reduce the dimensionality of the data and to simplify the optimization problem .
[ [ hub - language - characteristic . ] ] hub language characteristic .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + in the case of wikipedia , we observed that even though the training resources are scarce for certain language pairs , there often exists indirect training data . by considering a third language , which has training data with both languages in the pair
, we can use the composition of learned maps as a proxy .
we refer to this third language as a hub language .
a _ hub language _ is a language with a high proportion of non - empty documents in @xmath97 .
as we have mentioned , we only focus on multilingual documents that include at least two languages .
the prototypical example in the case of wikipedia is english .
our notion of the hub language could be interpreted in the following way .
if a non - english wikipedia page contains one or more links to variants of the page in other languages , english is very likely to be one of them .
that makes english a hub language .
we use the following notation to define subsets of the multilingual comparable corpus : let @xmath98 denote the index set of all multilingual documents with non - missing data for the @xmath18-th and @xmath21-th language : @xmath99 and let @xmath100 denote the index set of all multilingual documents with non missing data for the @xmath18-th language .
we now describe a two step approach to building a cross - lingual similarity matrix .
the first part is related to lsi and reduces the dimensionality of the data .
the second step refines the linear mappings and optimizes the linear dependence between data .
[ [ step-1-hub - language - based - dimensionality - reduction . ] ] step 1 : hub language based dimensionality reduction .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the first step in our method is to project @xmath101 to lower - dimensional spaces without destroying the cross - lingual structure . treating the nonzero columns of @xmath39 as observation vectors sampled from an underlying distribution @xmath102 , we can analyze the empirical cross - covariance matrices : @xmath103 where @xmath104 . by finding low - rank approximations of @xmath86 we can identify the subspaces of @xmath105 and @xmath106 that are relevant for extracting linear patterns between @xmath84 and @xmath85 .
let @xmath107 represent the hub language corpus matrix .
the lsi approach to finding the subspaces is to perform the singular value decomposition on the full @xmath108 covariance matrix composed of blocks @xmath86 . if @xmath109 is small for many language pairs ( as it is in the case of wikipedia ) , then many empirical estimates @xmath86 are unreliable , which can result in overfitting . for this reason , we perform the truncated singular value decomposition on the matrix @xmath110 \approx u s v^t$ ] , where @xmath111 .
we split the matrix @xmath112 vertically in blocks with @xmath113 rows : @xmath114^t$ ] .
note that columns of @xmath61 are orthogonal but columns in each @xmath105 are not ( columns of v are orthogonal ) .
let @xmath115 .
we proceed by reducing the dimensionality of each @xmath39 by setting : @xmath116 , where @xmath117 .
to summarize , the first step reduces the dimensionality of the data and is based on cl - lsi , but optimizes only the hub language related cross - covariance blocks .
[ [ step-2-simplifying - and - solving - sscor . ] ] step 2 : simplifying and solving sscor .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the second step involves solving a generalized version of canonical correlation analysis on the matrices @xmath118 in order to find the mappings @xmath64 .
the approach is based on the sum of squares of correlations formulation by kettenring @xcite , where we consider only correlations between pairs @xmath119 due to the hub language problem characteristic .
we will present the original unconstrained optimization problem , then a constrained formulation based on the hub language problem characteristic .
then we will simplify the constraints and reformulate the problem as an eigenvalue problem by using lagrange multipliers .
the original sum of squared correlations is formulated as an unconstrained problem : @xmath120 we solve a similar problem by restricting @xmath121 and omitting the optimization over non - hub language pairs .
let @xmath122 denote the empirical covariance of @xmath123 and @xmath124 denote the empirical cross - covariance computed based on @xmath123 and @xmath125 .
we solve the following constrained ( unit variance constraints ) optimization problem : @xmath126 the constraints @xmath127 can be simplified by using the cholesky decomposition @xmath128 and substitution : @xmath129 . by inverting the @xmath130 matrices and defining @xmath131 ,
the problem can be reformulated : @xmath132 a necessary condition for optimality is that the derivatives of the lagrangian vanish .
the lagrangian of ( [ squaredcorhub ] ) is expressed as : @xmath133 stationarity conditions give us : @xmath134 @xmath135 multiplying the equations ( [ dldxi ] ) with @xmath136 and applying the constraints , we can eliminate @xmath137 which gives us : @xmath138 plugging this into ( [ dldx1 ] ) , we obtain an eigenvalue problem : @xmath139 the eigenvectors of @xmath140 solve the problem for the first language . the solutions for @xmath141 are obtained from ( [ eqy1yi ] ) : @xmath142 .
note that the solution ( [ squaredcorhuboriginal ] ) can be recovered by : @xmath143 .
the linear transformation of the @xmath144 variables are thus expressed as : @xmath145 @xmath146 @xmath147 where @xmath42 is a diagonal matrix that normalizes @xmath148 , with @xmath149 .
[ [ remark . ] ] remark . + + + + + + + the technique is related to generalization of canonical correlation analysis ( gcca ) by carroll @xcite , where an unknown group configuration variable is defined and the objective is to maximize the sum of squared correlations between the group variable and the others .
the problem can be reformulated as an eigenvalue problem .
the difference lies in the fact that we set the unknown group configuration variable as the hub language , which simplifies the solution .
the complexity of our method is @xmath150 , where @xmath6 is the reduced dimension from the lsi preprocessing step , whereas solving the gcca method scales as @xmath151 , where @xmath33 is the number of samples ( see @xcite ) .
another issue with gcca is that it can not be directly applied to the case of missing documents . to summarize , we first reduced the dimensionality of our data to @xmath6-dimensional features and then found a new representation ( via linear transformation ) that maximizes directions of linear dependence between the languages
. the final projections that enable mappings to a common space are defined as : @xmath152
the main application on which we test the cross - lingual similarity is cross - lingual event linking . in online media streams
particularly news articles there is often duplication of reporting , different viewpoints or opinions , all centering around a single event .
the same events are covered by many articles and the question we address is how to find all the articles in different languages that are describing a single event . in this paper
we consider the problem of matching events from different languages .
we do not address the problem of detection of events and instead base our evaluation on an online system for detection of world events , event registry .
the events are represented by clusters of articles and so ultimately our problem reduces to finding suitable matchings between clusters with articles in different languages .
the problem of cross - lingual event linking is to match monolingual clusters of news articles that describing the same event across languages .
for example , we want to match a cluster of spanish news articles and a cluster of english news articles that both describe the same earthquake . each article @xmath153 is written in a language @xmath38 , where @xmath154 .
for each language @xmath38 , we obtain a set of monolingual clusters @xmath155 .
more precisely , the articles corresponding to each cluster @xmath156 are written in the language @xmath38 . given a pair of languages @xmath157 , @xmath158 and @xmath159 , we would like to identify all cluster pairs @xmath160 such that @xmath161 and @xmath162 describe the same event .
clusters composed of english and spanish news articles .
arrows link english articles with their spanish @xmath6-nearest neighbor matches according to the cross - lingual similarity.,scaledwidth=70.0% ] matching of clusters is a _ generalized matching _
we can not assume that there is only one cluster per language per event , nor can we assume complete coverage
i.e. , that there exists at least one cluster per event in every language .
this is partly due to news coverage which might be more granular in some languages , partly due to noise and errors in the event detection process .
this implies that we can not make assumptions on the matching ( e.g. , one - to - one or complete matching ) and excludes the use of standard weighted bipartite matching type of algorithms for this problem .
an example is shown in figure [ fig : clusters ] , where a cluster may contain articles which are closely matched with many clusters in a different language .
we are also seeking for an algorithm which does not do exhaustive comparison of all clusters , since that can become prohibitively expensive when working in a real - time setting .
more specifically , we wish to avoid testing cluster @xmath161 with all the clusters from all the other languages . performing exhaustive comparison would result in @xmath163 tests , where @xmath164 is the number of all clusters ( over all languages ) , which is not feasible when the number of clusters is on the order of tens of thousands .
we address this by testing only clusters that are connected with at least one @xmath6-nearest neighbor ( marked as _ candidate clusters _ in figure [ fig : clusters ] ) . in order to identify clusters that are equivalent to cluster @xmath161
, we have developed a two - stage algorithm . for a cluster @xmath161
, we first efficiently identify a small set of candidate clusters and then find those clusters among the candidates , which are equivalent to @xmath161 .
an example is shown in figure [ fig : clusters ] .
the details of the first step are described in algorithm [ cluster_merge_algo1 ] .
the algorithm begins by individually inspecting each news article @xmath165 in the cluster @xmath161 . using a chosen method for computing cross - lingual document similarity ( see section [ sec : models ] )
, it identifies the 10 most similar news articles to @xmath165 in each language @xmath166 . for each similar article @xmath167
, we identify its corresponding cluster @xmath162 and add it to the set of candidates .
the set of candidate clusters obtained in this way is several orders of magnitude smaller than the number of all clusters , and at most linear with respect to the number of news articles in cluster @xmath161 . in practice , clusters contain highly related articles and as such similar articles from other languages mostly fall in only a few candidate clusters .
although computed document similarities are approximate , our assumption is that articles in different languages describing the same event will generally have a higher similarity than articles about different events .
while this assumption does not always hold , redundancy in the data mitigates these false positives . since we compute the 10 most similar articles for each article in @xmath161 , we are likely to identify all the relevant candidates for cluster @xmath161 .
@xmath168 the second stage of the algorithm determines which ( if any ) of the candidate clusters are equivalent to @xmath161 .
we treat this task as a supervised learning problem . for each candidate cluster @xmath169
, we compute a vector of learning features that should be indicative of whether @xmath161 and @xmath162 are equivalent or not and apply a binary classification model that predicts if the clusters are equivalent or not .
the classification algorithm that we used to train a model was a linear support vector machine ( svm ) method @xcite .
we use three groups of features to describe cluster pair @xmath170 .
the first group is based on * cross - lingual article links * , which are derived using cross - lingual similarity : each news article @xmath165 is linked with its @xmath171-nearest neighbors articles from all other languages ( 10 per each language ) .
the group contains the following features : * ` linkcount ` is the number of times any of the news articles from @xmath162 is among @xmath171-nearest neighbors for articles from @xmath161 . in other words
, it is the number of times an article from @xmath161 has a very similar article ( i.e. , is among 10 most similar ) in @xmath162 . * ` avgsimscore ` is the average similarity score of the links , as identified for ` linkcount ` , between the two clusters .
the second group are * concept - related features*. articles that are imported into event registry are annotated by disambiguating mentioned _ entities _ and _ keywords _ to the corresponding wikipedia pages @xcite .
whenever barack obama is , for example , mentioned in the article , the article is annotated with a link to his wikipedia page . in the same way ,
all mentions of entities ( people , locations , organizations ) and ordinary keywords ( e.g. , bank , tax , ebola , plane , company ) are annotated .
although the spanish article about obama will be annotated with his spanish version of the wikipedia page , in many cases we can link the wikipedia pages to their english versions .
this can be done since wikipedia itself provides information regarding which pages in different languages represent the same concept / entity . using this approach ,
the word `` avin '' in a spanish article will be annotated with the same concept as the word `` plane '' in an english article .
although the articles are in different languages , the annotations can therefore provide a language - independent vocabulary that can be used to compare articles / clusters . by analyzing all the articles in clusters @xmath161 and @xmath162
, we can identify the most relevant entities and keywords for each cluster .
additionally , we can also assign weights to the concepts based on how frequently they occur in the articles in the cluster . from the list of relevant concepts and corresponding weights
, we consider the following features : * ` entitycossim ` is the cosine similarity between vectors of entities from clusters @xmath161 and @xmath162 . *
` keywordcossim ` is the cosine similarity between vectors of keywords from clusters @xmath161 and @xmath162 . * ` entityjaccardsim ` is jaccard similarity coefficient @xcite between sets of entities from clusters @xmath161 and @xmath162 . * ` keywordjaccardsim ` is jaccard similarity coefficient between sets of keywords from clusters @xmath161 and @xmath162 .
the last group of features contains three * miscellaneous features * that seem discriminative but are unrelated to the previous two groups : * ` hassamelocation ` feature is a boolean variable that is true when the location of the event in both clusters is the same .
the location of events is estimated by considering the locations mentioned in the articles that form a cluster and is provided by event registry . *
` timediff ` is the absolute difference in hours between the two events .
the publication time and date of the events is computed as the average publication time and date of all the articles and is provided by event registry . * ` shareddates ` is determined as the jaccard similarity coefficient between sets of date mentiones extracted from articles .
we use extracted mentions of dates provided by event registry , which uses an extensive set of regular expressions to detect and normalize mentions of dates in different forms .
we will describe the main dataset for building cross - lingual models which is based on wikipedia and then present two sets of experiments .
the first set of experiments establishes that the hub based approach can deal with language pairs where little or no training data is available .
the second set of experiments compares the main approaches that we presented on the task of mate retrieval and the task of event linking .
finally , we examine how different choices of features impact the event linking performance . to investigate the empirical performance of the low - rank approximations we will test the algorithms on a large - scale , real - world multilingual dataset that we extracted from wikipedia by using inter - language links for alignment .
this results in a large number of weakly comparable documents in more than @xmath172 languages .
wikipedia is a large source of multilingual data that is especially important for the languages for which no translation tools , multilingual dictionaries as eurovoc @xcite , or strongly aligned multilingual corpora as europarl @xcite are available .
documents in different languages are related with so called inter - language links that can be found on the left of the wikipedia page .
the wikipedia is constantly growing .
there are currently 12 wikipedias with more than 1 million articles , @xmath173 with more than 100k articles , @xmath174 with more than 10k articles , and @xmath175 with more than @xmath176 articles .
each wikipedia page is embedded in the page tag .
first , we check if the title of the page starts with a wikipedia namespace ( which includes categories and discussion pages ) and do not process the page if it does .
then , we check if this is a redirection page and we store the redirect link because inter - language links can point to redirection links also . if none of the above applies , we extract the text and parse the wikipedia markup .
currently , all the markup is removed .
we get inter - language link matrix using previously stored redirection links and inter - language links .
if an inter - language link points to the redirection we replace it with the redirection target link .
it turns out that we obtain the matrix @xmath177 that is not symmetric , consequently the underlying graph is not symmetric .
that means that existence of the inter - language link in one way ( i.e. , english to german ) does not guarantee that there is an inter - language link in the reverse direction ( german to english ) . to correct this
we transform this matrix to be symmetric by computing @xmath178 and obtaining an undirected graph . in the rare case that after symmetrization we have multiple links pointing from the document , we pick the first one that we encountered .
this matrix enables us to build an alignment across all wikipedia languages ( for dumps available in 2013 ) . in this subsection
, we will investigate the empirical performance of hub cca approach .
we will demonstrate that this approach can be successfully applied even in the case of fully missing alignment information . to this purpose , we select a subset of wikipedia languages containing three major languages , english ( 4,212k articles)_en _ ( hub language ) , spanish ( 9,686k articles)_es _ , russian ( 9,662k articles)_ru _ , and five minority ( in terms of wikipedia sizes ) languages , slovenian ( 136k articles)_sl _ , piedmontese ( 59k articles)_pms _ , waray - waray ( 112k articles)_war _ ( all with about 2 million native speakers ) , creole ( 54k articles)_ht _ ( 8 million native speakers ) , and hindi ( 97k articles)_hi _ ( 180 million native speakers ) . for preprocessing , we remove the documents that contain less than 20 different words ( referred to as stubs ) and remove words occurring in less than 50 documents as well as the top 100 most frequent words ( in each language separately ) .
we represent the documents as normalized tfidf @xcite weighted vectors .
the idf scores are computed for each language based on its aligned documents with the english wikipedia .
the english language idf scores are based on all english documents for which aligned spanish documents exist .
the evaluation is based on splitting the data into training and test sets .
we select the test set documents as all multilingual documents with at least one nonempty alignment from the list : ( _ hi _ , _ ht _ ) , ( _ hi _ , _ pms _ ) , ( _ war _ , _ ht _ ) , ( _ war _ , _ pms _ ) .
this guarantees that we cover all the languages .
moreover this test set is suitable for testing the retrieval through the hub as the chosen pairs have empty alignments .
the remaining documents are used for training . in table
[ table : train_test ] , we display the corresponding sizes of training and test documents for each language pair . on the training set , we perform the two step procedure to obtain the common document representation as a set of mappings @xmath64 .
a test set for each language pair , @xmath179 , consists of comparable document pairs ( linked wikipedia pages ) , where @xmath180 is the test set size .
we evaluate the representation by measuring mate retrieval quality on the test sets : for each @xmath38 , we rank the projected documents @xmath181 according to their similarity with @xmath182 and compute the rank of the mate document @xmath183 . the final retrieval score ( between -100 and 100 ) is computed as : @xmath184 . a score that is less than 0 means that the method performs worse than random retrieval and a score of 100 indicates perfect mate retrieval .
the mate retrieval results are included in table [ table : retrieval ] .
we observe that the method performs well on all pairs of languages , where at least 50,000 training documents are available(_en _ , _ es _ , _ ru _ , _ sl _ ) .
we note that taking @xmath185 or @xmath186 multilingual topics usually results in similar performance , with some notable exceptions : in the case of ( _ ht _ , _ war _ ) the additional topics result in an increase in performance , as opposed to ( _ ht _ , _ pms _ ) where performance drops , which suggests overfitting .
the languages where the method performs poorly are _
ht _ and _ war _ , which can be explained by the quality of data ( see table [ table : rank ] and explanation that follows ) . in case of _ pms
_ , we demonstrate that solid performance can be achieved for language pairs ( _ pms _ , _ sl _ ) and ( _ pms _ , _ hi _ ) , where only 2,000 training documents are shared between _ pms _ and _ sl _ and no training documents are available between _ pms _ and _ hi_. also observe that in the case of ( _ pms _ , _ ht _ ) the method still obtains a score of 62 , even though training set intersection is zero and _ ht _ data is corrupted , which we will show in the next paragraph .
c|c|c|c|c|c|c|c|c| & en & es & ru & sl & hi & war & ht & pms + en & 671 - 4.64 & 463 - 4.29 & 369 - 3.19 & 50.3 - 2 & 14.4 - 2.76 & 8.58 - 2.41 & 17 - 2.32 & 16.6 - 2.67 + es & & 463 - 4.29 & 187 - 2.94 & 28.2 - 1.96 & 8.72 - 2.48 & 6.88 - 2.4 & 13.2 - 2 & 13.8 - 2.58 + ru & & 369 - 3.19 & 29.6 - 1.92 & 9.16 - 2.68 & 2.92 - 1.1 & 3.23 - 2.2 & 10.2 - 1.29 + sl & & 50.3 - 2 & 3.83 - 1.65 & 1.23 - 0.986 & 0.949 - 1.23 & 1.85 - 0.988 + hi & & 14.4 - 2.76 & 0.579 - 0.76 & 0.0 - 2.08 & 0.0 - 0.796 + war & & 8.58 - 2.41 & 0.043 - 0.534 & 0.0 - 1.97 + ht & & 17 - 2.32 & 0.0 - 0.355 + pms & & 16.6 - 2.67 + en & & 98 - 98 & 95 - 97 & 97 - 98 & 82 - 84 & 76 - 74 & 53 - 55 & 96 - 97 + es & 97 - 98 & & 94 - 96 & 97 - 98 & 85 - 84 & 76 - 77 & 56 - 57 & 96 - 96 + ru & 96 - 97 & 94 - 95 & & 97 - 97 & 81 - 82 & 73 - 74 & 55 - 56 & 96 - 96 + sl & 96 - 97 & 95 - 95 & 95 - 95 & & 91 - 91 & 68 - 68 & 59 - 69 & 93 - 93 + hi & 81 - 82 & 82 - 81 & 80 - 80 & 91 - 91 & & 68 - 67 & 50 - 55 & 87 - 86 + war & 68 - 63 & 71 - 68 & 72 - 71 & 68 - 68 & 66 - 62 & & 28 - 48 & 24 - 21 + ht & 52 - 58 & 63 - 66 & 66 - 62 & 61 - 71 & 44 - 55 & 16 - 50 & & 62 - 49 + pms & 95 - 96 & 96 - 96 & 94 - 94 & 93 - 93 & 85 - 85 & 23 - 26 & 66 - 54 & + we further inspect the properties of the training sets by roughly estimating the fraction @xmath187 for each english training matrix and its corresponding mate matrix , where @xmath188 and @xmath189 denote the number of rows and columns respectively .
the denominator represents the theoretically highest possible rank the matrix @xmath0 could have .
ideally , these two fractions should be approximately the same - both aligned spaces should have reasonably similar dimensionality .
we display these numbers as pairs in table [ table : rank ] .
ht & en pms + 0.81 0.89 & 0.8 0.89 & 0.98 0.96 & 1 1 & 0.74 0.56 & 1 0.22 & 0.89 0.38 + it is clear that in the case of the creole language only at most @xmath190 documents are unique and suitable for the training .
though we removed the stub documents , many of the remaining documents are nearly the same , as the quality of some smaller wikipedias is low .
this was confirmed for the creole , waray - waray , and piedmontese languages by manual inspection .
the low quality documents correspond to templates about the year , person , town , etc . and contain very few unique words .
there is also a problem with the quality of the test data .
for example ,
if we look at the test pair ( _ war _ , _ ht _ ) only 386/534 waray - waray test documents are unique but on the other side almost all creole test documents ( 523/534 ) are unique .
this indicates a poor alignment which leads to poor performance . in order to determine
how accurately we can predict cluster equivalence , we performed two experiments in a multilingual setting using english , german and spanish languages for which we had labelled data to evaluate the linking performance . in the first experiment , we tested how well the individual approaches for cross - lingual article linking perform when used for linking the clusters about the same event .
in the second experiment we tested how accurate the prediction model is when trained on different subsets of learning features . to evaluate the prediction accuracy for a given dataset we used 10-fold cross validation .
we created a manually labelled dataset in order to evaluate cross - lingual event linking using two human annotators .
the annotators were provided with an interface listing the articles , their content from and top concepts for a pair of clusters and their task was to determine if the clusters were equivalent or not ( i.e. , discuss same event ) . to obtain a pair of clusters @xmath170 to annotate , we first randomly chose a cluster @xmath161 , used algorithm [ cluster_merge_algo1 ] to compute a set of potentially equivalent clusters @xmath56 and randomly chose a cluster @xmath169 .
the dataset provided by the annotators contains 808 examples , of which 402 are equivalent clusters pairs and 406 are not .
clusters in each learning example are either in english , spanish or german . although event registry imports articles in other languages as well , we restricted our experiments to these three languages .
we chose only these three languages since they have very large number of articles and clusters per day which makes the cluster linking problem hard due to large number of possible links . in section
[ sec : models ] , we described three main algorithms for identifying similar articles in different languages .
these algorithms were @xmath6-means , lsi and hub cca .
as a training set , we used common wikipedia alignment for all three languages . to test which of these algorithms performed best , we made the following test .
for each of the three algorithms , we analyzed all articles in event registry and for each article computed the most similar articles in other languages . to test
how informative the identified similar articles are for cluster linking we then trained three classifiers as described in section [ algo : features ]
one for each algorithm .
each classifier was allowed to use as learning features * only * the cross - lingual article linking features for which values are determined based on the selected algorithm ( @xmath6-means , lsi and hub cca ) .
the results of the trained models are shown in table [ table : linkingevalalgos ] .
we also show how the number of topics ( the dimensions of the latent space ) influences the quality , except in the case of the @xmath6-means algorithm , where only the performance on 500 topic vectors is reported , due to higher computational cost .
we observe that , for the task of cluster linking , lsi and hub cca perform comparably and both outperform @xmath6-means .
we also compared the proposed approaches on the task of wikipedia mate retrieval ( the same task as in section [ experiments : hubcca ] ) .
we computed the average ( over language pairs ) mean reciprocal rank ( amrr ) @xcite performance of the different approaches on the wikipedia data by holding out @xmath191 aligned test documents and using @xmath192 aligned documents as the training set .
figure [ pic : amrr ] shows amrr score as the function of the number of feature vectors .
it is clear that hub cca outperforms lsi approach and @xmath6-means lags far behind when testing on wikipedia data .
the hub cca approach with @xmath193 topic vectors manages to perform comparably to the @xmath194-based approach with @xmath176 topic vectors , which shows that the @xmath195 method can improve both model memory footprint as well as similarity computation time .
furthermore , we inspected how the number of topics influences the accuracy of cluster linking . as we can see from table [ table : linkingevalalgos ] choosing a number of features larger than @xmath193 barely affects linking performance , which is in contrast with the fact that additional topics helped to improve ammr , see figure [ pic : amrr ] .
such differences may have arisen due to different domains of training and testing ( wikipedia pages versus news articles ) .
we also analyzed how cluster size influences the accuracy of cluster linking .
we would expect that if the tested pair of clusters has a larger number of articles then the classifier should be able to more accurately predict whether the clusters should be linked or not .
the reasoning is that the large clusters would provide more document linking information ( more articles mean more links to other similar articles ) as well as more accurately aggregated semantic information . in the case of smaller clusters ,
the errors of the similarity models have greater impact which should decrease the performance of the classifier , too .
to validate this hypothesis we have split the learning examples into two datasets one containing cluster pairs where the combined number of articles from both clusters is below 20 and one dataset where the combined number is 20 or more .
the results of the experiment can be seen in table [ table : linkingevalalgoslargesmall ] .
as it can be seen , the results confirm our expectations : for smaller clusters it is indeed harder to correctly predict if the cluster pair should be merged or not .
the hub cca attains higher precision and classification accuracy on the task of linking small cluster pairs than the other methods , while lsi is slightly better on linking large cluster pairs .
the gain in precision of lsi over hub cca on linking large clusters is much smaller than the gain in precision of hub cca over lsi on linking small clusters .
for that reason we decided to use hub cca as the similarity computation component in our system .
table[row sep = crcr]100 0.574094231082746 + 200 0.645978450670063 + 300 0.682226506762222 + 400 0.710722094692085 + 500 0.731517169423203 + 600 0.746223870375363 + 700 0.758312886611436 + 800 0.766279310559898 + 900 0.768159613081689 + 1000 0.76200517888255 + ; ; table[row sep = crcr]100 0.388773162433835 + 200 0.49061571487968 + 300 0.550801867794634 + 400 0.592565604079171 + 500 0.625657099289601 + 600 0.65240747960386 + 700 0.675602522126668 + 800 0.694149727717174 + 900 0.711511053167843 + 1000 0.726455461653153 + ; ; table[row sep = crcr]100 0.203799120889507 + 200 0.275537657758881 + 300 0.334738107118325 + 400 0.378380649062479 + 500 0.409668545430712 + 600 0.427765004442194 + 700 0.454449272479224 + 800 0.477291178317808 + 900 0.490692600634559
+ 1000 0.518392823243787 + ; ; @xmath196 % + hub cca & 78.2/79.6/80.3 & 76.3/78.0/80.5 & 81.6/82.1/79.9 & 78.9/80.0/80.2 + lsi & 78.9/78.7/80.6 & 76.8/77.0/78.7 & 83.3/80.6/83.6 & 79.9/78.8/81.1 + @xmath6-means & 73.9/-/- & 69.5/-/- & 84.6/-/- & 76.3/-/- + @xmath196 % + hub cca & 81.2 - 77.8 & 80.5 - 74.5 & 91.3 - 57.5 & 85.6 - 64.9 + lsi & 82.8 - 76.4 & 81.3 - 70.9 & 93.1 - 57.5 & 86.8 - 63.5 + @xmath6-means & 75.5 - 71.2 & 72.8 - 70.8 & 95.3 - 36.2 & 82.5 - 47.9 + in the second experiment , we evaluate how relevant individual groups of features are to correctly determine cluster equivalence . for this purpose , we computed accuracy using individual groups of features , as well as using different combination of groups . since hub cca had the best performance of the three algorithms , we used it to compute the values of the cross - lingual article linking features .
the results of the evaluation are shown in table [ table : linkingeval ] .
we can see that using a single group of features , the highest prediction accuracy can be achieved using concept - related features .
the classification accuracy in this case is 88.5% . by additionally including also the cross - lingual article linking features , the classification accuracy rises slightly to 89.4% . using all three groups of features , the achieved accuracy is 89.2% . to test
if the accuracy of the predictions is language dependent we have also performed the evaluations separately on individual language pairs .
for this experiment we have split the annotated learning examples into three datasets , where each dataset contained only examples for one language pair . when training the classifier all three groups of features were available .
the results are shown in table [ table : langpaireval ] .
we can see that the performance of cluster linking on the english - german dataset is the highest in terms of accuracy , precision , recall and @xmath196 .
the performance on the english - spanish dataset is comparable to the performance on the english - german dataset , where the former achieves higher recall ( and slightly higher @xmath196 score ) , while the latter achieves higher precision .
a possible explanation of these results is that the higher quantity and quality of english - german language resources leads to a more accurate cross - lingual article similarity measure as well as to a more extensive semantic annotation of the articles .
based on the performed experiments , we can make the following conclusions . the cross - lingual similarity algorithms provide valuable information that can be used to identify clusters that describe the same event in different languages .
for the task of cluster linking , the cross - lingual article linking features are however significantly less informative compared to the concept - related features that are extracted from the semantic annotations .
nevertheless , the cross - lingual article similarity features are very important for two reasons .
the first is that they allow us to identify for a given cluster a limited set of candidate clusters that are potentially equivalent .
this is a very important feature since it reduces the search space by several orders of magnitude .
the second reason these features are important is that concept annotations are not available for all articles as the annotation of news articles is computationally intensive and can only be done for a subset of collected articles .
the prediction accuracies for individual language pairs are comparable although it seems that the achievable accuracy correlates with the amount of available language resources .
@xmath196 % + hub cca & @xmath197 & @xmath198 & @xmath199 & @xmath200 + concepts & @xmath201 & @xmath202 & @xmath203 & @xmath204 + misc & @xmath205 & @xmath206 & @xmath207 & @xmath208 + hub cca + concepts & @xmath209 & @xmath210 & @xmath211 & @xmath212 + hub cca + misc & @xmath213 & @xmath214 & @xmath215 & @xmath216 + concepts + misc & @xmath217 & @xmath218 & @xmath219 & @xmath220 + all & @xmath221 & @xmath222 & @xmath223 & @xmath224 + @xmath196 % + en , de & @xmath225 & @xmath226 & @xmath227 & @xmath228 + en , es & @xmath229 & @xmath230 & @xmath231 & @xmath232 + es , de & @xmath233 & @xmath234 & @xmath235 & @xmath236 + one of the main advantages of our approach is that it is highly scalable .
it is fast , very robust to quality of training data , easily extendable , simple to implement and has relatively small hardware requirements .
the similarity pipeline is the most computationally intensive part and currently runs on a machine with two intel xeon e5 - 2667 v2 , 3.30ghz processors with 256 gb of ram .
this is sufficient to do similarity computation over a large number of languages if needed .
it currently uses wikipedia as a freely available knowledge base and experiments show that the similarity pipeline dramatically reduces the search space when linking clusters . currently , we compute similarities over @xmath237 languages with tags : _ eng _ , _ spa _ , _ deu _ , _ zho _ , _ ita _ , _ fra _ , _ rus _ , _ swe _ , _ nld _ , _ tur _ , _ jpn _ , _ por _ , _ ara _ , _ fin _ , _ ron _ , _ kor _ , _ hrv _ , _ tam _ , _ hun _ , _ slv _ , _ pol _ , _ srp _ , _ cat _ , _ ukr _ but we support any language from the top @xmath238 wikipedia languages .
our data stream is newsfeed ( http://newsfeed.ijs.si/ ) which provides 430k unique articles per day .
our system currently computes 2 million similarities per second , that means that we compute @xmath239 similarities per day .
we store one day buffer for each language which requires 1.5 gb of memory with documents stored as 500-dimensional vectors .
we note that the time complexity of the similarity computations scales linearly with dimension of the feature space and does not depend on number of languages . for each article
, we compute the top @xmath171 most similar ones in every other language . for all linear algebra matrix and vector operations ,
we use high performance numerical linear algebra libraries as blas , openblas and intel mkl , which currently allows us to process more than one million articles per day . in our current implementation
, we use the variation of the hub approach .
our projector matrices are of size @xmath240 so every projector takes about @xmath241 gb of ram .
moreover , we need proxy matrices of size @xmath242 for every language pair . that is 0.5 gb for @xmath237 languages and @xmath243 gb for @xmath238 languages . all together
we need around 135 gb of ram for the system with 100 languages .
usage of proxy matrices enables the projection of all input documents in the common space and handling language pairs with missing or low alignment .
that enables us to do block - wise similarity computations further improving system efficiency .
our code can therefore be easily parallelized using matrix multiplication rather than performing more matrix - vector multiplications .
this speeds up our code by a factor of around @xmath244 in this way , we obtain some caching gains and ability to use vectorization . our system is also easily extendable .
adding a new language requires the computation of a projector matrix and proxy matrices with all other already available languages .
in this paper we have presented a cross - lingual system for linking events in different languages .
building on an existing system , event registry , we present and evaluate several approaches to compute a cross - lingual similarity function . we also present an approach to link events and evaluate effectiveness of various features .
the final pipeline is scalable both in terms of number of articles and number of languages , while accurately linking events . on the task of mate retrieval
, we observe that refining the lsi - based projections with hub cca leads to improved retrieval precision , but the methods perform comparably on the task of event linking .
further inspection showed that the cca - based approach reached a higher precision on smaller clusters .
the interpretation is that the linking features are highly aggregated for large clusters , which compensates the lower per - document precision of lsi .
another possible reason is that the advantage that we show on wikipedia is lost on the news domain .
this hypothesis could be validated by testing the approach on documents from a different domain .
the experiments show that the hub cca - based features present a good baseline , which can greatly benefit from additional semantic - based features .
even though in our experiments the addition of cca - based features to semantic features did not lead to great performance improvements , there are two important benefits in the approach .
first , the linking process can be sped up by using a smaller set of candidate clusters .
second , the approach is robust to languages where semantic extraction is not available , due to scarce linguistic resources .
currently the system is loosely - coupled the language component is built independently from the rest of the system , in particular the linking component .
it is possible that better embeddings can be obtained by methods that jointly optimize a classification task and the embedding .
another point of interest is to evaluate the system on languages with scarce linguistic resources , where semantic annotation might not be available . for this purpose
, the labelled dataset of linked clusters should be extended first .
the mate retrieval evaluation showed that even for language pairs with no training set overlap , the hub cca recovers some signal . in order to further improve the performance of the classifier for cluster linking
, additional features should also be extracted from articles and clusters and checked if they can increase the accuracy of the classification . since the amount of linguistic resources vary significantly from language to language it would also make sense to build a separate classifier for each language pair . intuitively , this should improve performance since weights of individual learning features could be adapted to the tested pair of languages .
janez brank , gregor leban , and marko grobelnik . a high - performance multithreaded approach for clustering a stream of documents . in _ proceedings of the 17th international multiconference information society 2014 , volume e , ljubljana , slovenia _ , pages 58 , 2014 .
dumais , t.a .
letsche , m.l .
littman , and t.k landauer .
automatic cross - language retrieval using latent semantic indexing . in _
aaai spring symposium on cross - language text and speech retrieval .
american association for artificial intelligence , vol . 16 .
, page 21 , 1997 .
philipp koehn , hieu hoang , alexandra birch , chris callison - burch , marcello federico , nicola bertoldi , brooke cowan , wade shen , christine moran , richard zens , chris dyer , ondej bojar , alexandra constantin , and evan herbst .
moses : open source toolkit for statistical machine translation . in _ proceedings of the 45th annual meeting of the acl on interactive poster and demonstration sessions _ , acl 07 , pages 177180 , prague , czech republic , 2007 .
association for computational linguistics .
gregor leban , blaz fortuna , janez brank , and marko grobelnik .
cross - lingual detection of world events from news articles . in _ proceedings of the 13th international semantic web conference _ , pages 2124 , riva del garda - trentino , italy , 2014 .
gregor leban , blaz fortuna , janez brank , and marko grobelnik .
event registry : learning about world events from news . in _ proceedings of the companion publication of the 23rd international conference on world wide web companion _ ,
www companion 14 , pages 107110 , seoul , republic of korea , 2014 . international world wide web conferences steering committee .
kalev leetaru and philip a schrodt .
gdelt : global data on events , location , and tone , 19792012 . in _
international studies association ( isa ) annual convention _ , volume 2 , page 4 , san francisco , california , usa , 2013 .
david milne and ian h. witten . learning to link with wikipedia . in _ proceedings of the 17th acm conference on information and knowledge management _ , cikm 08 , pages 509518 , napa valley , california , usa , 2008 .
david mimno , hanna m. wallach , jason naradowsky , david a. smith , and andrew mccallum .
polylingual topic models . in _ proceedings of the 2009 conference on empirical methods in natural language processing : volume 2 - volume 2 _ , emnlp 09 , pages 880889 , stroudsburg , pa , usa , 2009 .
association for computational linguistics .
andrej muhi , jan rupnik , and primo kraba .
cross - lingual document similarity . in _
information technology interfaces ( iti ) , proceedings of the iti 2012 34th international conference on _ , pages 387392 , cavtat / dubrovnik , croatia , 2012 .
ieee .
john c platt , kristina toutanova , and wen - tau yih .
translingual document representations from discriminative projections . in _ proceedings of the 2010 conference on empirical methods in natural language processing _ ,
pages 251261 , massachusetts , usa , 2010 .
association for computational linguistics .
martin potthast , benno stein , and maik anderka . a wikipedia - based multilingual retrieval model . in _ advances in information retrieval ,
30th european conference on information retrieval research ( ecir ) _ , pages 522530 ,
glasgow , uk , 2008 .
bruno pouliquen , ralf steinberger , and olivier deguernel .
story tracking : linking similar news over time and across languages . in _ proceedings of the workshop on multi - source multilingual information extraction and summarization _ , pages 4956 , manchester , united kingdom , 2008 .
association for computational linguistics .
jose mara lvarez rodrguez , emilio rubiera azcona , and luis polo paredes . promoting government controlled vocabularies for the semantic web : the eurovoc thesaurus and the cpv product classification system .
, page 111 , 2008 .
jan rupnik , andrej muhic , and primoz skraba .
multilingual document retrieval through hub languages . in _ proceedings of the 15th multiconference on information society 2012 ( is-2012 ) _ , pages 201204 , ljubljana , slovenia , 2012 .
duo zhang , qiaozhu mei , and chengxiang zhai .
cross - lingual latent topic extraction . in _ proceedings of the 48th annual meeting of the association for computational linguistics _ , pages 11281137 , uppsala , sweden , 2010 .
association for computational linguistics .
lei zhang and achim rettinger .
semantic annotation , analysis and comparison : a multilingual and cross - lingual text analytics toolkit . in _ proceedings of the demonstrations at the 14th conference of the european chapter of the association for computational linguistics ( eacl 2014 ) _ , pages 1316 ,
gothenburg , sweden , april 2014 .
association for computational linguistics . | in today s world , we follow news which is distributed globally .
significant events are reported by different sources and in different languages . in this work ,
we address the problem of tracking of events in a large multilingual stream . within a recently developed system event registry @xcite
we examine two aspects of this problem : how to compare articles in different languages and how to link collections of articles in different languages which refer to the same event . taking a multilingual stream and clusters of articles from each language
, we compare different cross - lingual document similarity measures based on wikipedia .
this allows us to compute the similarity of any two articles regardless of language .
building on previous work , we show there are methods which scale well and can compute a meaningful similarity between articles from languages with little or no direct overlap in the training data . using this capability
, we then propose an approach to link clusters of articles across languages which represent the same event .
we provide an extensive evaluation of the system as a whole , as well as an evaluation of the quality and robustness of the similarity measure and the linking algorithm . |
a _ quandle _ is an algebraic structure whose axioms are transcriptions of the three reidemeister moves , making quandles a natural source of invariants of knots and links . in @xcite , joyce associated a finitely generated quandle to each tame knot or link @xmath0 and proved that this knot quandle determines the knot type up to orientation - reversing homeomorphism of pairs @xmath1 .
since then , much work has been done in the study of knot invariants defined in terms of the knot quandle , such as @xcite , @xcite , etc .
in particular , for any finite quandle @xmath2 there are invariants of knots and links defined in terms of the set @xmath3 of quandle homomorphisms from the knot quandle @xmath4 into a finite target quandle @xmath2 .
since isomorphic target quandles @xmath2 and @xmath5 define the same invariants , we would like to have a convenient way of distinguishing finite quandles .
moreover , understanding the structure of these finite quandles can lead to a better understanding of their link invariants . in this paper
we describe a two - variable polynomial invariant of finite quandles .
this polynomial encodes a set with multiplicities arising from counting trivial actions of elements on other elements ; the polynomial is a convenient way to encode this information with the added benefit of obtaining integer - valued invariants via specialization of @xmath6 and @xmath7 . while not a complete invariant , this polynomial does seem to be quite good at distinguishing finite quandles
indeed , in some cases the invariant completely determines the structure of the quandle .
one motivation for this construction comes from consideration of the nature of identity elements and kernels of homomorphisms .
specifically , groups have a unique identity element which acts trivially on the rest of the group , and the preimage of this element under a homomorphism measures how much information about the domain group is lost in the homomorphism . unlike groups , quandles do not have a unique identity element ; rather , every element in a quandle acts trivially on itself and possibly some other elements .
we might say that in a group , the `` trivial action '' is concentrated in a single element , while in a quandle it is distributed throughout the set .
our polynomial may then be understood as a way of quantifying how this `` trivial action '' is distributed .
the paper is organized as follows . in section 2 ,
we recall the definition of a quandle and some examples of quandles .
we then define the @xmath8 polynomial and list some examples . in section 3
, we define a generalization of @xmath8 for subquandles of a larger quandle and a @xmath8-based laurent polynomial associated to a homomorphism between finite quandles .
we the use @xmath8 to define a family of link invariants which are jazzed - up versions of the familiar quandle counting invariant . in section 4
, we list some questions for future research and end with a table listing the @xmath8 values for all quandles with order up to 5 .
we begin with a definition from joyce , @xcite .
the axioms can be understood as transcriptions of the three reidemeister moves where each arc in the knot diagram corresponds to a quandle element and the quandle operation corresponds to an arc going through a crossing .
some well - known examples of quandles include : @xmath9 @xmath10 let @xmath11 be a finite quandle .
we can describe the algebraic structure of @xmath12 with an @xmath13 matrix @xmath14 satisfying @xmath15=k$ ] where @xmath16 .
in other words , the matrix of @xmath12 is just the quandle operation table without the `` @xmath17 ' 's .
this quandle matrix notation is convenient for symbolic computation purposes @xcite , @xcite . [ qp ] @xmath18 that is
, @xmath19 is the number of elements of @xmath12 which act trivially on @xmath17 , while @xmath20 is the number of elements of @xmath12 on which @xmath17 acts trivially .
these numbers are easily computed from the matrix of @xmath12 , simply by going through rows and columns and counting occurrences of the row number .
note that if @xmath12 has infinite cardinality , then @xmath20 and @xmath19 may not be finite and @xmath21 may not be defined .
[ iso ] if @xmath12 and @xmath22 are isomorphic finite quandles , then @xmath23 .
suppose @xmath24 is an isomorphism of quandles and fix @xmath25 .
then for every @xmath26 , we have @xmath27 , so @xmath28 and we have @xmath29 . repeating the same argument with @xmath30 yields the opposite inequality , and we have @xmath31 .
a similar argument shows that @xmath32 .
then we have @xmath33 @xmath34 \quad \mathrm{and } \quad qp_{t_n}(s , t ) = ns^nt^n.\ ] ] evaluating @xmath35 yields @xmath36 , the cardinality of @xmath12
. @xmath37 thus , we may regard @xmath21 as a kind of enhanced version of cardinality for finite quandles .
it s easy to see that @xmath21 completely determines the quandle structure for some quandles , such as @xmath38 above .
in fact , @xmath8 distinguishes all quandles of order 3 and 4 and all non - latin quandles of order 5 , as shown in table 1 .
there are examples of non - isomorphic quandles with the same @xmath8 values , as the next example shows .
the @xmath8 polynomial distinguishes between some quandles which have the same orbit decomposition but different structure maps ( see @xcite ) .
@xmath39 \quad \mathrm{and } \quad m_{q'}=\left[\begin{array}{cccccc } 1 & 3 & 2 & 1 & 3 & 2 \\ 3 & 2 & 1 & 3 & 2 & 1 \\ 2 & 1 & 3 & 2 & 1 & 3 \\ 4 & 6 & 5 & 4 & 6 & 5 \\ 6 & 5 & 4 & 6 & 5 & 4 \\ 5 & 4 & 6 & 5 & 4 & 6 \end{array}\right]\ ] ] a latin quandle @xmath12 satisfies @xmath40 [ lat ] let @xmath12 be a latin quandle and fix an element @xmath41 . since @xmath12 is a quandle , the @xmath42 entry of @xmath14 is @xmath43 , so @xmath44 and @xmath45 . now , suppose @xmath46 .
then there is a column @xmath47 with @xmath15=i$ ] .
but then row @xmath43 is not a permutation of @xmath48 , contradicting the fact that @xmath12 is latin .
hence , we must have @xmath49 . similarly , if @xmath50 then there is some @xmath47 such that @xmath51=j$ ] , and then we have @xmath52 , contradicting our conclusion above .
hence @xmath53 .
since @xmath43 was arbitrary , we have @xmath54 direct computations show that the converse of proposition [ lat ] has no counterexamples with cardinality less than 7 .
thus we have : every quandle @xmath12 with @xmath55 is latin .
we may also define the @xmath8 invariant for non - quandle racks ; such racks may have @xmath56 , as in the next example .
@xmath57 @xmath58 indeed , if @xmath59 for a rack @xmath60 , then @xmath60 is not a quandle .
in this section we define a generalization of @xmath21 which gives information about how a subquandle is embedded in a larger quandle and we use this to define a family of link invariants .
@xmath61 if a finite quandle @xmath62 is a disjoint union of several subquandles ( e.g. , the orbit subquandles ) then we have @xmath63 @xmath64 [ ker ] if @xmath24 is injective , then every exponent in @xmath65 is nonnegative ; if @xmath66 is surjective , then every exponent in @xmath65 is nonpositive . if @xmath66 is an isomorphism , then @xmath67 .
this is similar to the proof of proposition [ iso ] .
thus , @xmath65 has a philosophical similarity to the kernel of a homomorphism .
note , however , that the converse of proposition [ ker ] is not true , as the next example shows .
recall that for any knot or link @xmath68 , there is an associated _ knot quandle _ @xmath4 , and that for a given finite quandle @xmath2 the set of quandle homomorphisms @xmath69 is a source of computable knot invariants .
specifically , we can take the cardinality of the set , which gives us an integer - valued invariant .
alternatively , we can count the homomorphisms weighted by a cocycle in one of the various quandle cohomology theories ( described in @xcite , @xcite etc . )
; these cocycles provide a way of squeezing extra information out of the set of homomorphisms .
we can use the subquandle polynomials of the image of each homomorphism in a similar way .
@xmath70 normally , we encode sets with multiplicities whose elements are integers as polynomials where the multiplicities appear as coefficients and the elements of the set are powers of a variable ; in this case , however , the elements of the set are already polynomials , and polynomials with polynomial powers seem a little awkward
. however , we can derive convenient polynomial - valued specializations of @xmath71 by choosing values of @xmath6 and @xmath7 in @xmath72 and evaluating .
@xmath73 @xmath74 in general , specializations of @xmath71 contain more information than the unadorned counting invariant , as the next example shows .
@xmath75.$}\ ] ]
in this section we list a few questions whose answers may prove interesting . under
what conditions are two finite quandles @xmath8-equivalent ?
what information in addition to @xmath21 must be specified in order to fully determine the quandle type of @xmath12 up to isomorphism ? for which polynomials @xmath76 $ ] is the set of quandles @xmath12 with @xmath77 nonempty ?
it s easy to see that such a @xmath78 must have all positive coefficients whose sum is @xmath36 and each term must have degree @xmath79 satisfying @xmath80 ( or @xmath81 in the rack case ) , but what other conditions are necessary or sufficient
? can we deduce from a quandle s polynomial whether the quandle is isomorphic to a conjugation quandle , an alexander quandle , a symplectic quandle , or a direct product , semidirect product , or an abelian extension of these ?
for which specialized quandle polynomial invariants is there a quandle 2-cocycle ( see @xcite ) such that @xmath82 is equal to the quandle cocycle invariant ?
do we gain any information by combining the two , perhaps by setting @xmath83 where @xmath84 is the sum of the cocycle values at the crossings in @xmath68 ( see @xcite ) ? | we define a two - variable polynomial invariant of finite quandles . in many cases
this invariant completely determines the algebraic structure of the quandle up to isomorphism .
we use this polynomial to define a family of link invariants which generalize the quandle counting invariant .
keywords : finite quandles , link invariants 2000 msc : 57m27 , 176d99 |
spectral collocation method is a classical and powerful tool to solve integral and differential equations .
suppose that the equation is defined on a finite interval @xmath4 $ ] , the basic idea of this approach is to approximate the solution of the equation by its polynomial interpolant of the form @xmath5 where @xmath6 is a set of distinct nodes and @xmath7 are the lagrange fundamental polynomials .
the values @xmath8 are determined by requiring that the interpolant @xmath9 satisfies the equation exactly at the nodes @xmath10 . to ensure rapid convergence of the spectral collocation method , the interpolation nodes @xmath11 with the distribution of density @xmath12 are preferable , and the ideal candidates are the zeros or extrema of classical orthogonal polynomials such as gegenbauer polynomials , or more generally , the jacobi polynomials ; cf .
@xcite . the interpolation procedure described above
is also known as spectral interpolation .
as is well known , the accuracy of spectral interpolation depends on the regularity of the underlying function @xmath13 , with exponential rate if @xmath13 is analytic in a neighborhood containing the interval @xmath4 $ ] ; we refer to @xcite for relevant results and @xcite for fast implementation . to this end , it is worthwhile to recall that the starting point of these proofs is the so - called hermite integral formula .
more precisely , let @xmath14 be the bernstein ellipse : @xmath15 the bernstein ellipse @xmath14 has the foci at @xmath16 with the major and minor semi - axes given by @xmath17 and @xmath18 , respectively .
suppose that @xmath13 is analytic on and within @xmath14 for some @xmath19 , it follows from the hermite integral formula ( * ? ? ?
* theorem 3.6.1 ) that @xmath20 where @xmath21 with @xmath22 being a positive normalization constant .
this in turn implies that @xmath23 \\ z \in \mathcal{e}_{\rho } } } \left| \frac { \omega_n(x ) } { \omega_n(z ) } \right|,\end{aligned}\ ] ] where @xmath24 , @xmath25 denotes the length of the circumference of @xmath14 , and @xmath26 is the distance from @xmath14 to the interval @xmath4 $ ] . for polynomial interpolation at the jacobi points ,
i.e. , the nodes @xmath27 are the roots of @xmath28-th jacobi polynomial @xmath29 ( @xmath30 ) and the polynomial @xmath31 in is taken to be @xmath32 , the properties of @xmath29 on the bernstein ellipse are essential in the analysis of convergence rate . for the gegenbauer polynomials @xmath33 ( corresponding to the special cases @xmath34 of the jacobi polynomials ) , by noting that ( see ( * ? ? ? * lemma 3.1 ) ) @xmath35 where @xmath36 the following asymptotic estimate of gegenbauer polynomials on the bernstein ellipse is obtained by xie , wang and zhao in ( * ? ? ?
* theorem 3.2 ) : there exists @xmath37 such that @xmath38 for @xmath19 , @xmath39 and @xmath40 , where @xmath41 the estimate plays an important role in the rigorous proofs of exponential convergence of gegenbauer interpolation and spectral differentiation conducted in @xcite .
later , in a paper regarding superconvergence of jacobi - gauss type spectral interpolation @xcite , wang , zhao and zhang have made use of the following estimate of the lower bound for jacobi polynomial on the bernstein ellipse : @xmath42 where @xmath43 is regular for @xmath44 , and @xmath45 but @xmath46 ; see ( * ? ? ? * equation ( 4.7 ) ) .
this result follows directly from the asymptotic formula of jacobi polynomials ( * ? ? ?
* theorem 8.21.9 ) .
we note that , however , except for the very special cases like @xmath47 , the explicit form of @xmath48 is not available .
we also note that there is a close connection between polynomial interpolation and the potential theory ( * ? ? ?
* chapter 5 ) . more specifically , let us define the discrete potential function associated with the nodes @xmath49 by @xmath50 this function is harmonic in the complex plane except at @xmath49 and can be viewed as the potential generated by all @xmath49 if each @xmath51 is interpreted as a point charge of strength @xmath52 and the repulsion is inverse - linear .
clearly , we have @xmath53 .
thus , if we choose @xmath51 to be the zeros of @xmath0 , then the extrema of @xmath54 implies the extrema of the corresponding potential @xmath55 as well .
it is the aim of the present research to conduct more complete studies of jacobi polynomials on the bernstein ellipse , including the explicit formula , extrema of the absolute value and the asymptotic estimate .
our main contributions are listed below : * an explicit formula of @xmath0 is derived in the variable of parametrization , which generalizes valid for the gegenbauer case .
* the extrema of @xmath1 on the bernstein ellipse @xmath14 are identified under some assumptions on the parameters .
we show that the maximum value is attained at one of the endpoints of the major axis if @xmath2 .
this particularly extends ( * ? ? ?
* theorem 4.5.1 ) established by ismail , which is valid for the gegenbauer polynomials . for the minimum value
, we are able to show that for a large class of gegenbauer polynomials , it is attained at two endpoints of the minor axis . *
we provide a more refined and computable asymptotic estimate as well as a lower bound for the jacobi polynomials on the bernstein ellipse , which generalizes concerning the gegenbauer polynomials .
the rest of this paper is organized as follows .
we first give a brief review of jacobi polynomials in section [ sec : property ] , which includes some basic properties that will be used later .
section [ sec : jacobi ellipse ] is devoted to the explicit representation of jacobi polynomials on the bernstein ellipse .
the extrema of jacobi polynomials on the bernstein ellipse are discussed in section [ sec : jacobiextrema ] .
the identification of maximum value relies on a three - term recurrence relation for the coefficients arising in the explicit formula . for the minimum value
, we first deal with the chebyshev polynomials of the first and second kinds and then extend the results to gegenbauer polynomials . the asymptotic estimate and the lower bound of jacobi polynomials on the bernstein ellipse
are presented in section [ sec : jacobibound ] .
in this section , we collect some basic properties of jacobi polynomials which will be used in the subsequent analysis .
all these properties can be found in the classical book of szeg @xcite .
let @xmath29 denote the jacobi polynomial of degree @xmath28 , which is defined explicitly by @xmath56 the jacobi polynomials are orthogonal over @xmath4 $ ] with respect to the weight function @xmath57 , that is , @xmath58 where @xmath59 is the kronecker delta and @xmath60 from the explicit formula , it is easily seen that @xmath61 where the leading coefficient @xmath62 is given by @xmath63 let @xmath64 with @xmath30 .
the maximum of @xmath65 on the interval @xmath4 $ ] is given by ( see ( * ? ? ? * theorem 7.32.1 ) ) @xmath66}\left|p_{n}^{(\alpha,\beta)}(x)\right| = \left\{\begin{array}{cc } { \displaystyle \binom{n+q}{n } } , & \mbox{if $ q \geq -\frac{1}{2}$ } , \\ [ 15pt ] { |p_{n}^{(\alpha,\beta)}(\tilde{x})| } , & \mbox{if $ q < -\frac{1}{2}$ } , \end{array } \right.\end{aligned}\ ] ] where @xmath67 is one of the two maximum points nearest @xmath68 .
indeed , when @xmath69 , the maximum of @xmath65 is attained at one of the endpoints @xmath70 .
when @xmath3 , the jacobi polynomials are , up to some positive constants , also known as gegenbauer ( or the ultraspherical ) polynomials @xmath71 .
more precisely , we have @xmath72 the orthogonality of gegenbauer polynomials reads @xmath73 where @xmath74 since the weight function @xmath75 is an even function , it is readily seen that following symmetry relations hold : @xmath76 thus , @xmath77 is an even function for even @xmath28 and an odd function for odd @xmath28 .
the chebyshev polynomials of the first and second kinds are @xmath78 respectively .
when @xmath79 , the chebyshev polynomials have the following simple representations in the variable of parametrization : @xmath80 they are special cases of gegenbauer polynomials , and the relations are given by @xmath81 equivalently , one has @xmath82 finally , let @xmath83 be the zeros of gegenbauer polynomials . by ( * ? ? ?
* theorem 6.21.1 ) , it follows that @xmath84 where @xmath85 denotes the integer part of @xmath86 .
thus , the positive zeros of a gegenbauer polynomial @xmath71 strictly decrease with respect to the parameter @xmath87 .
it is the aim of this section to prove the following theorem , which gives an explicit representation of @xmath0 on the bernstein ellipse in the variable of parametrization .
[ thm : jacobi ellipse ] for @xmath88 , i.e. , @xmath89 we have @xmath90 where the coefficients are given by @xmath91,\end{gathered}\ ] ] for @xmath92 . to show theorem [ thm : jacobi ellipse ]
, we note that chebyshev polynomials of the first kind @xmath93 has a simple explicit formula on the bernstein ellipse , the strategy is then to expand jacobi polynomials in terms of @xmath93 .
the connection formula between two different families of jacobi polynomials are stated in the following lemma ( see ( * ? ? ?
* theorem 7.1.1 ) ) .
[ lem : jacobi connection ] assume that @xmath94 then the connection coefficients are given by @xmath95,\end{aligned}\ ] ] where @xmath96 is the pochhammer symbol and @xmath97 = \sum_{k=0}^{\infty } \frac { ( a_1)_k ( a_2)_k ( a_3)_k } { ( b_1)_k ( b_2)_k } \frac{z^k}{k!}\end{aligned}\ ] ] is the generalized hypergeometric function . with the aid of lemma [ lem : jacobi connection ] , we are now ready to prove theorem [ thm : jacobi ellipse ] . [ [ proof - of - theorem - thmjacobi - ellipse ] ] proof of theorem [ thm : jacobi ellipse ] + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + note that @xmath102 has a simple expression @xmath103 on @xmath14 . substituting this formula into the last equation of gives us the desired result .
this completes the proof of theorem [ thm : jacobi ellipse ] .
suppose that @xmath104 ( i.e. , @xmath105 ) , we obtain from the following trigonometric representations of jacobi polynomials @xmath106 the above formula seems to be new , except for the special case @xmath3 ( cf .
* formula ( 4.9.19 ) ) ) .
[ rk : ultraspherical ] when @xmath107 , the coefficients @xmath108 can be further simplified with the help of the properties of hypergeometric function @xmath109 . indeed , on account of the fact ( see ( * ? ? ?
* theorem 3.5.5 ) ) that @xmath110=\frac{\gamma\left(\frac{1}{2}\right)\gamma\left(c+\frac{1}{2}\right)\gamma\left(\frac{a+b+1}{2}\right)\gamma\left(c-\frac{a+b-1}{2}\right ) } { \gamma\left(\frac{a+1}{2}\right)\gamma\left(\frac{b+1}{2}\right)\gamma\left(c-\frac{a-1}{2}\right)\gamma\left(c-\frac{b-1}{2}\right)},\ ] ] it follows from and straightforward calculations that @xmath111 { 0 } , & \mbox{if $ n - k$ is odd}. \end{array } \right.\end{aligned}\ ] ] this particularly implies that @xmath112 up to some constant factors , this recovers which was derived via the three - term recurrence relation of gegenbauer polynomials in @xcite .
the approach used therein , however , seems difficult to be generalized to handle the jacobi case .
in this section , we will consider the extrema of jacobi polynomials on the bernstein ellipse . the maximum value and the minimum value will be discussed in subsections [ sec : max ] and [ sec : min ] , respectively . by
* theorem 4.5.1 ) , it is known that for the gegenbauer polynomials @xmath71 , @xmath113 , @xmath88 with @xmath114 is attained at the right endpoint of the major axis .
it comes out that similar property holds for the jacobi polynomials @xmath115 with @xmath116 , which is our main result of this section .
[ prop : location ] for @xmath117 and @xmath118 , we have 1 . if @xmath119 and @xmath120 , then the maximum value of @xmath121 is attained uniquely at the right endpoint of the major axis , i.e. , @xmath122 2 . if @xmath123 and @xmath120 , then the maximum value of @xmath121 is attained uniquely at the left endpoint of the major axis , i.e. , @xmath124 3 .
if @xmath125 , then the maximum value of @xmath121 is attained at two endpoints of the major axis , i.e. , @xmath126 moreover , the maximum value can only be attained at these two real points @xmath127 if @xmath128 .
the assertion in item @xmath129 corresponds to the case of gegenbauer polynomials mentioned at the very beginning .
moreover , the condition @xmath116 implies that @xmath130 . by setting @xmath105 in the above theorem
, we recover the result concerning maximum of @xmath65 over the orthogonal interval @xmath4 $ ] , as explained after .
the proof of theorem [ prop : location ] relies on the explicit formula of @xmath32 on the bernstein ellipse established in theorem [ thm : jacobi ellipse ] .
the essential issue here is to determine the signs of the coefficients @xmath108 appearing in under various conditions on the parameters @xmath131 and @xmath132 ; see proposition [ prop : sign ] below .
to proceed , we start with the following proposition which reveals a recurrence relation for the coefficients @xmath133 and plays a fundamental role in the sequel .
[ prop : recurr ] with @xmath108 defined in , we have , for each @xmath134 and @xmath135 , @xmath136 with initial conditions @xmath137 this completes the proof of proposition [ prop : recurr ] .
from , it is readily seen that if @xmath3 , the three - term recurrence relation can be simplified as @xmath149 in addition , note that the coefficients @xmath108 in involve the hypergeometric function @xmath109 , it would be helpful to use the recurrence relation in actual computations . as a consequence of proposition [ prop : recurr ] , we are able to determine the signs of the coefficients @xmath150 in the following proposition . [ prop : sign ] for @xmath151 , we have 1 . if @xmath119 and @xmath120 , then @xmath152 .
2 . if @xmath123 and @xmath120 , then @xmath153 .
3 . if @xmath154 , then @xmath155 = 0 , & \hbox{if $
n - k$ is odd . }
\end{array } \right.\ ] ] if @xmath47 , then @xmath156 = 0 , & \hbox{if $ k=1,2,\ldots , n-1$. } \end{array } \right.\ ] ] if @xmath119 and @xmath120 , it is easily seen that @xmath157 and @xmath158 for @xmath159 .
these , together with the recurrence relation and the fact that both of the initial values @xmath160 and @xmath161 are positive ( see ) , imply the assertion in item @xmath162 . similarly , if @xmath123 and @xmath120 , we have @xmath163 and @xmath164 for @xmath159 . since @xmath165 and @xmath166 in this case , we again obtain from that @xmath167 is an alternating sequence , as required . finally , if @xmath168 , the assertion in item @xmath129 follows immediately from .
this completes the proof of proposition [ prop : sign ] .
we are now ready to prove theorem [ prop : location ] .
[ [ proof - of - theorem - proplocation ] ] proof of theorem [ prop : location ] + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + if @xmath119 and @xmath120 , we recall from item @xmath162 in proposition [ prop : sign ] that @xmath152 for @xmath151 . on account of and , it is straightforward to see that @xmath169 thus , @xmath170 can be achieved if and only if @xmath171 , which is .
similarly , if @xmath123 and @xmath120 , a combination of item @xmath172 in proposition [ prop : sign ] and implies that @xmath173 hence , the maximum value can be achieved if and only if @xmath174 , as shown in . to show , we see from remark [ rk : ultraspherical ] , and that @xmath175 thus , the maximum value can be achieved when @xmath176 and follows .
moreover , if @xmath154 , since @xmath177 is strictly positive ( see ) for @xmath151 , the maximum value can only be achieved at two endpoints of the major axis .
this completes the proof of theorem [ prop : location ] . for the very special case @xmath47 , the jacobi polynomials ( up to a normalization constant ) are the chebyshev polynomials of the first kind @xmath93
; see . in this case
, we obtain from that for @xmath88 , @xmath178 it is then clear that @xmath179 is attained if and only if @xmath180 , i.e. , at @xmath181 points @xmath182 the identification of minimum value of @xmath121 on the bernstein ellipse @xmath14 is , in general , much more involved than that of its maximum value . since the minimum value will be zero when @xmath105 , we will restrict our attention to @xmath19 and focus on the ultraspherical case @xmath3 in this section . in view of the relations and ,
the results will be presented in terms of @xmath93 , @xmath183 and @xmath71 . to provide some intuition about the location where gegenbauer polynomials attain the minimum value , we perform some numerical experiments of @xmath184 with @xmath88 ; see figures [ fig : gegenbauer1][fig : gegenbauer3 ] for different choices of the parameters @xmath87 , @xmath28 and @xmath185 .
note that we omit the numerical results for @xmath186 and even @xmath187 , since they are similar to those shown in figure [ fig : gegenbauer3 ] .
the numerical studies imply that the minimum value depends on the parameters @xmath87 , @xmath28 , @xmath185 , and further suggest the following conjecture concerning the observations .
it is conjectured that 1 .
if @xmath188 and @xmath189 is odd , @xmath190 is attained at @xmath191 for @xmath19 , i.e. , at two endpoints of the minor axis .
if @xmath188 and @xmath192 is even , there exists a critical value @xmath193 depending on the parameters @xmath28 and @xmath87 such that @xmath190 is attained at @xmath191 for @xmath194 .
if @xmath186 and @xmath192 , there exists a critical value @xmath195 depending on the parameters @xmath28 and @xmath87 such that @xmath190 is attained at @xmath196 for @xmath197 , i.e. , at two endpoints of the major axis . in what follows , we shall prove items @xmath162 and @xmath172 of the above conjecture under the assumptions that @xmath198 and @xmath199 ( for item @xmath172 ) .
with @xmath200 for @xmath201 ( left ) , @xmath202 ( middle ) and @xmath203 ( right ) .
here @xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath201 ( left ) , @xmath202 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath201 ( left ) , @xmath202 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] with @xmath200 for @xmath207 ( left ) , @xmath208 ( middle ) and @xmath203 ( right ) . here
@xmath204 ranges from @xmath205 to @xmath206 .
, title="fig:",width=181 ] we first deal with the chebyshev polynomials of the first and second kinds .
the relevant results , on one hand , will provide important insights for the general case , on the other hand they are crucial in further analysis .
[ thm : cheby1kind ] for @xmath19 and @xmath209 , we have @xmath210 and the minimum value is attained at @xmath181 points @xmath211 from it is readily seen that @xmath212 attains its minimum value if and only if @xmath213 i.e. , for @xmath214 with @xmath215 .
the desired results follow immediately .
[ rk : chebodd ] from , it follows that , for odd @xmath28 , the minimum value of @xmath216 can be attained at @xmath217 .
we next proceed to the chebyshev polynomial of the second kind @xmath218 .
the following lower bound can be found in ( * ? ? ?
* formula ( 1.53 ) ) : @xmath219 our next theorem shows that this lower bound is attainable only when @xmath28 is odd .
if @xmath28 is even , a new and attainable lower bound will be presented under some conditions on @xmath185 .
[ thm : chebyumin ] for @xmath118 , we have @xmath220 { \displaystyle \frac{\rho^{n+1 } + \rho^{-n-1}}{\rho + \rho^{-1 } } } , & \hbox{if $ n$ is even and $ \rho\geq\rho_n^*$ , } \end{array } \right.\end{aligned}\ ] ] where @xmath221 is the unique root of the equation @xmath222 and where @xmath223 moreover , in both cases the the minimum value is attained if and only if @xmath224 , i.e. , at two endpoints of the minor axis .
the above theorem can actually be seen from a remarkable connection between @xmath218 and the kernel @xmath225 arising in the contour integral representation of the remainder term of an @xmath28-point gauss quadrature for the chebyshev weight function of the second kind .
more precisely , let @xmath226 be an analytic function on and within the bernstein ellipse @xmath14 .
the gaussian quadrature rule for the chebyshev weight function of the second kind reads @xmath227 where @xmath228 are the zeros of the chebyshev polynomial of the second kind @xmath218 , and @xmath229 are the corresponding christoffel numbers .
the remainder term @xmath230 admits the following contour representation : @xmath231 where the kernel @xmath225 is given by @xmath232 where @xmath233 and the second equality follows from ( * ? ? ?
* equation 3.613.3 ) .
the above formula particularly implies that @xmath234 is proportional to the reciprocal of @xmath235 if @xmath88 . since the maximum value of @xmath236 over @xmath14 has been studied in the context of estimating the remainder term @xmath237 in @xcite , theorem [ thm : chebyumin ] follows directly from (
* theorem 5.2 ) and ( * ? ? ?
* theorem 1 ) . for completeness
, we include a more direct proof in what follows .
[ [ proof - of - theorem - thmchebyumin ] ] proof of theorem [ thm : chebyumin ] + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + by , it is readily seen that @xmath238 denote by @xmath239 the function appearing on the right hand side of .
it is then equivalent to consider the minimum of @xmath239 for @xmath240 $ ] .
we start with the easy case that @xmath28 is odd . by
, it is clear that @xmath241 and the lower bound on the right hand side is attained if and only if @xmath242 and @xmath243 , which gives @xmath244 or @xmath245 .
thus , the minimum can only be attained at two endpoints of the minor axis , i.e. , at the points @xmath246 , as desired .
if @xmath28 is even , a straightforward calculation shows that @xmath247 \nonumber \\ & = \frac{2(\cos\theta)^2}{a_2(\rho ) - \cos(2\theta ) } \left [ \left ( \frac{a_{n+1}(\rho)}{a_1(\rho ) } \right)^2 - \left ( \frac{\cos((n+1)\theta)}{\cos\theta } \right)^2 \right].\end{aligned}\ ] ] to this end , we note that , on one hand , @xmath248 where @xmath249 .
it then follows from the property of @xmath218 that @xmath250 } \left| \frac{\cos(n+1)\theta}{\cos\theta } \right| = n+1,\end{aligned}\ ] ] and the upper bound can be attained if and only if @xmath244 or @xmath251 .
on the other hand , it is easily seen that the function @xmath252 is strictly increasing for @xmath253 and @xmath28 fixed .
hence , if @xmath254 , we see from and that @xmath255 in addition , since @xmath256 the second case in @xmath257 follows . it is also easy to see that the minimum is attained if and
only if @xmath244 or @xmath251 .
this completes the proof of theorem [ thm : chebyumin ] .
[ rem : rhon ] for even @xmath28 and @xmath258 , one can conclude from ( * ? ? ?
* theorem 1 ) that the minimum value of @xmath259 is attained at some @xmath260 with @xmath261 , which is slightly off the imaginary axis . moreover , from ( * ? ? ?
* theorem 2 ) we know that @xmath262 is a strictly decreasing sequence and @xmath263 as @xmath264 ; see figure [ fig : rho ] for an illustration . for @xmath265.,title="fig:",width=264 ]
we finally come to the gegenbauer polynomials @xmath71 . besides the trivial case@xmath266 , we have the following theorem .
[ thm : gegenbauer ] let @xmath267 be the unique root of with @xmath268 . for @xmath269 and @xmath187 ,
the minimum value of @xmath270 is attained at two endpoints of the minor axis , i.e. , @xmath271 provided @xmath272 , or @xmath273 and @xmath28 is odd .
we precede the proof of theorem [ thm : gegenbauer ] with the following lemma .
[ lemma : minr ] let @xmath88 and define @xmath274 where @xmath275 .
then , for @xmath276 and @xmath277 , @xmath278 see ( * ? ? ?
* lemma 4.1 ) .
[ [ proof - of - theorem - thmgegenbauer ] ] proof of theorem [ thm : gegenbauer ] + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + let @xmath279 be the zeros of @xmath71 arranged in decreasing order .
the symmetry relation implies that @xmath280 where @xmath281 is the leading coefficient of @xmath71 .
moreover , we see that @xmath282 for @xmath283 .
let @xmath284 be the positive zeros of the chebyshev polynomials of the second kind @xmath218 .
again , we could rewrite @xmath218 as @xmath285 by , it follows that @xmath286 for @xmath283 . to find @xmath190 is equivalent to find @xmath287 . a combination of and gives @xmath288 in view of , lemma [ lemma : minr ] , theorem [ thm : chebyumin ] and the monotonicity of @xmath289 aforementioned in remark [ rem : rhon ] , we conclude that all the terms on the right hand side of attain their maximum values at @xmath246 for @xmath290 .
therefore , @xmath291 attains its minimum at two endpoints of the minor axis provided @xmath272 .
the case for @xmath273 and odd @xmath28 can be proved in a similar manner .
we only need to replace @xmath218 in by the chebyshev polynomials of the first kind @xmath216 , and to make use of remark [ rk : chebodd ] instead .
the details are left to the interested readers .
this completes the proof of theorem [ thm : gegenbauer ] .
from the explicit formula of gegenbauer polynomials on the bernstein ellipse , the authors in @xcite derive an asymptotic estimate of @xmath33 as shown in . due to the complexity of the coefficients
@xmath108 given in , it is difficult to apply the same approach to obtain the asymptotic estimate of jacobi polynomials on the bernstein ellipse . to this end
, we note that a more computable form has been given in @xcite , where the authors actually consider asymptotics of polynomials orthogonal with respect to a modified jacobi weight function @xmath292 with @xmath30 and @xmath293 being real analytic and strictly positive on @xmath4 $ ] .
based on the riemann - hilbert ( rh ) approach @xcite , various asymptotics of the monic / orthonomal polynomials in the complex plane have been derived in @xcite , which in particular includes a full asymptotic expansion for the monic polynomials outside @xmath4 $ ] .
to state the relevant results , we need the function @xmath294,\ ] ] where @xmath295 is analytic in @xmath296 $ ] and behaves like @xmath297 as @xmath298 .
this function is a conforming mapping from @xmath296 $ ] onto the exterior of the unit circle .
thus , @xmath299.\ ] ] as in @xcite , we also define the szeg function of a weight @xmath300 by @xmath301,\ ] ] and @xmath302 the function @xmath303 is analytic for @xmath304 $ ] .
let @xmath305 denote the monic orthogonal polynomial of degree @xmath28 associated with .
it is shown in ( * ? ? ?
* theorem 1.4 ) that @xmath306 has an asymptotic expansion of the form @xmath307 , \quad n\rightarrow\infty,\end{aligned}\ ] ] uniformly valid for @xmath297 in any compact subsets of @xmath296 $ ] .
the functions @xmath308 , which are analytic on @xmath309 $ ] , are rational in @xmath310 .
they are explicitly computable via the rh approach but with more complicated form as @xmath311 increases .
the first two terms are @xmath312 and @xmath313.\end{aligned}\ ] ] for an efficient numerical calculations of the higher - order terms @xmath308 , we refer to recent work @xcite .
obviously , the classical jacobi polynomials correspond to the case @xmath314 in .
we then have the following asymptotic estimate of jacobi polynomials on the bernstein ellipse in the variable of parametrization .
[ prop : asyellip ] for @xmath88 , i.e. , @xmath315 we have , for large @xmath28 , @xmath316 where @xmath317 and @xmath318 furthermore , the error is uniformly bounded for @xmath319 with @xmath19 .
we first derive the uniform asymptotics of @xmath32 for @xmath320 $ ] . in view of the facts that @xmath321 and
@xmath322 it then follows from and that @xmath323 , \quad n\rightarrow\infty,\end{aligned}\ ] ] uniformly valid for @xmath297 in any compact subsets of @xmath296 $ ] , where @xmath62 is defined as in . using asymptotic formulas for the gamma functions ( see ( * ? ? ?
* formulas 5.11.3 and 5.11.13 ) ) , we deduce that @xmath324.\ ] ] this , together with , implies that @xmath325 , \quad n\rightarrow\infty,\end{gathered}\ ] ] where @xmath326 is given in . if @xmath327 $ ] , which can be parameterized through the argument @xmath328 as shown in , it is straightforward to check that @xmath329 a combination of the above two formulas and gives @xmath330,\end{gathered}\ ] ] where @xmath331 is defined as in and the asymptotics is valid uniformly for @xmath88 with @xmath19 . by using the above uniform asymptotic , it is straightforward to derive the desired result and this completes the proof of proposition [ prop : asyellip ] .
one should compare the asymptotic estimate with .
it is worthwhile to point out that the error in is of order @xmath332 . indeed
, a full asymptotic expansion of @xmath32 in terms of powers of @xmath52 on the bernstein ellipse @xmath14 can be derived by combining and a full asymptotic expansion of the leading coefficient @xmath62 .
we note that this form of asymptotic expansion has been mentioned in ( * ? ? ?
* theorem 8.21.9 ) , but without explicit formulas for the coefficients .
as a direct consequence of proposition [ prop : asyellip ] , we have @xmath333 where @xmath334 and @xmath44
. a further application of proposition [ prop : asyellip ] is that we are able to derive the following lower bound for the jacobi polynomial on the bernstein ellipse , which particularly implies a more explicit expression of the constant @xmath48 appearing in . for @xmath335
, we have @xmath336 where @xmath337 is a positive constant and @xmath338 for large @xmath28 .
moreover , @xmath339 ( 1-\rho^{-2})^{\alpha+\frac{1}{2 } } , & \mbox{if $ -1 < \alpha=\beta < -\frac{1}{2}$. } \end{array } \right.\ ] ] the lower bound follows immediately from proposition [ prop : asyellip ] and the elementary inequality @xmath340 , @xmath19 and @xmath341 , it is easily seen that @xmath342 hence , if @xmath343 , @xmath344 attains its maximum value at @xmath345 and its minimum value at @xmath346 , which gives us .
in this paper , we have investigated several basic properties of jacobi polynomials on the bernstein ellipse , which include the explicit formula , extrema of the absolute values as well as a refined asymptotic estimate .
these results provide some further insight into jacobi polynomials and can be adaptable to some practical applications such as establishing an explicit error bound of the spectral interpolation at the jacobi nodes .
haiyong wang thanks the hospitality of the school of mathematical sciences at fudan university where the present research was initiated .
his work was supported by the national natural science foundation of china under grant 11671160 .
lun zhang was partially supported by the program for professor of special appointment ( eastern scholar ) at shanghai institutions of higher learning ( no .
shh1411007 ) , by national natural science foundation of china ( no . 11501120 ) and by grant ezh1411513 from fudan university .
f. w. j. olver , d. w. lozier , r. f. boisvert , and c. w. clark , editors . .
department of commerce , national institute of standards and technology , washington , dc ; cambridge university press , cambridge , 2010 . with 1 cd - rom ( windows , macintosh and unix ) . | in this paper , we are concerned with jacobi polynomials @xmath0 on the bernstein ellipse with motivation mainly coming from recent studies of convergence rate of spectral interpolation .
an explicit representation of @xmath0 is derived in the variable of parametrization .
this formula further allows us to show that the maximum value of @xmath1 over the bernstein ellipse is attained at one of the endpoints of the major axis if @xmath2 . for the minimum value ,
we are able to show that for a large class of gegenbauer polynomials ( i.e. , @xmath3 ) , it is attained at two endpoints of the minor axis .
these results particularly extend those previously known only for some special cases .
moreover , we obtain a more refined asymptotic estimate for jacobi polynomials on the bernstein ellipse .
* keywords : * spectral method , jacobi polynomials , bernstein ellipse , extrema , asymptotic estimate * ams classifications : * 65n35 , 65d05 , 41a05 , 41a25 . |
not only can strong gravitational lensing yield impressive images , from multiply imaged quasars over deformed galaxies to partial or full einstein rings , it is also an invaluable tool to estimate the mass of the deflector .
moreover , since the precise positions and deformations of the images of a source depend on the exact distribution of the mass , gravitational lensing even holds the promise of constraining the shape of said mass distribution . in principle
, this strong lens inversion sounds fairly straightforward : use a particular model and optimize its parameters so that it reproduces the observations as well as possible . in practice however , one is hindered by gravitational lensing degeneracies which allow a wide range of mass distributions to produce the exact same image configuration .
on the level of parametric lens inversion results , this can manifest itself as parameter degeneracies ( e.g. @xcite ) , but at their core the degeneracies are or course present at the level of the mass distribution . there , the mass sheet degeneracy ( e.g. @xcite or @xcite ) will rescale the projected mass distribution while adding a constant density mass sheet ( or disc ) , only affecting the time delays between the images .
recently it was shown that a similar procedure is still possible when multiple sources at different redshifts are observed @xcite .
while this degeneracy therefore can still be broken when time delay measurements are present , the situation is far worse with respect to the so - called monopole degeneracy @xcite .
this type of degeneracy can be used to redistribute the mass in between the images , without changing any of the observable properties of the images . by presenting a further generalization of this mass sheet degeneracy
, this article will illustrate an important relation between different degenerate versions of a lensing mass distribution .
this will illustrate how degeneracies are closely related to the substructure in the mass density , which , in turn , offers insights into what one may constrain about the lensing mass .
after briefly introducing the lensing formalism and used notations in section [ sec : formalism ] , the basics of the monopole and mass sheet degeneracies are reviewed in section [ sec : basics ] .
this will serve as the starting point for some extensions in section [ sec : extensions ] after which an experiment in section [ sec : experiment ] will illustrate precisely how difficult it can be to estimate the parameters of a specific type of mass distribution , an elliptical generalization of a navarro - frenk - white profile @xcite .
finally , in section [ sec : conclusions ] we shall discuss what the implications of these degeneracies are and what different types of information can help constrain .
in this section , the same notation as @xcite is used , which the interested reader can consult for a thorough review . the lens equation ( [ eq : lenseq ] ) in essence
states that due to the gravitational deflection @xmath0 of a light ray , when looking in a direction @xmath1 one sees what would be observed in direction @xmath2 if the lens effect could be turned off : @xmath3 the deflection angle @xmath0 depends on the two - dimensional projected mass distribution @xmath4 in the following fashion : @xmath5 in these equations , the geometry of the lensing scenario is given by the angular diameter distances @xmath6 , @xmath7 and @xmath8 , describing the distance to the lens , to the source and the distance between lens and source respectively
. it will often be convenient to use the so - called critical density : @xmath9 which depends on the redshift of a source through the angular diameter distances involved . in the special case of a circularly symmetric mass distribution , for convenience assumed to be centered on the origin of the coordinate system
, the expression of the deflection angle reduces to : @xmath10 in which @xmath11 represents the mass enclosed within radius @xmath12 : @xmath13 the deflection angle can be shown to be proportional to the gradient of the so - called lens potential @xmath14 which is related to the time light takes to travel from the source at position @xmath2 to an image at position @xmath1 : @xmath15 in this last equation , @xmath16 is the redshift of the gravitational lens .
since , in practice one is only interested in the difference in time delay between images of the same source , the constant in this expression is of no importance .
the curvature of the lens potential can be shown to represent the mass density at the corresponding point : @xmath17 finally , the magnification @xmath18 that an image at position @xmath1 experiences , can be calculated from the lens equation as follows : @xmath19 i.e. it is the jacobian determinant of this equation .
the so - called monopole degeneracy , first mentioned in @xcite and studied in more detail in @xcite , is both the least complex and most elusive degeneracy one can think of .
consider the mass distribution shown in the inset in fig .
[ fig : monopole ] .
being a circularly symmetric mass distribution , its lens effect is governed by equation ( [ eq : alphasymm ] ) . since this particular mass distribution
was constructed to have zero enclosed mass outside a particular radius , the same equation shows that it will not cause a deflection angle outside said radius .
this kind of mass density will be referred to as a monopole , since it can be described by only the monopole term in a multipole expansion .
an example will illustrate the importance of such a monopole . the elliptical mass distribution shown in the left panel of fig .
[ fig : monopole ] transforms a single source into five distinct images , as can be seen in the left panel of fig .
[ fig : monopoleeffect ] . indicated in the same figure , is a circular region in which none of the images are located , and which has the precise same diameter as the aforementioned monopole .
when adding the monopole in this location , this implies that the lens equation will not be affected outside the circular region ; only within will the deflection of light be affected .
that this is indeed the case , can clearly be seen in the right panel of fig .
[ fig : monopoleeffect ] .
there , the lens effect is calculated for the same source , but for the combined mass distribution shown in the right panel of fig .
[ fig : monopole ] . as was expected ,
the exact same five images are generated by this clearly different mass density .
the fact that the lens equation was modified inside the circular region can be seen in the change in critical lines in this region .
since the deflection is different inside the circular area , in principle it is even possible that extra images are generated there .
however , this particular example shows that this does not necessarily need to be the case , even for a rather large modification of the mass distribution .
this example already shows that by ` borrowing ' some mass from a relatively large region , it is easy to introduce ( or remove ) a mass peak in a mass density , without changing the images that are generated .
as was explicitly shown in @xcite however , it is also possible to use such monopoles as basis functions which allows a complex redistribution of the mass by adjusting their weights . as long as no basis function overlaps with an image
, none of the images will be affected by this operation .
some care needs to be taken not to introduce additional , unobserved images though . from the discussion above ,
it is already trivially clear that this type of degeneracy does not modify the image positions generated by a particular source .
the fact that the deflection angle is zero outside a certain radius , by equation ( [ eq : magnification ] ) implies that the magnification at the location of the images will not be affected either .
equation ( [ eq : lenspotential ] ) on the other hand shows that the lens potential outside said radius may change by a constant value . however , when measuring time delays between images , such a constant offset in the potential will have no effect , as can be seen in equation ( [ eq : timedelay ] ) .
the mass - sheet degeneracy is undoubtedly one of the most famous degeneracies in gravitational lensing .
contrary to what the name may suggest , to obtain a degenerate solution it is not sufficient to simply add a sheet of mass to an existing mass distribution .
instead , one has to rescale the original mass distribution as well , prompting the use of the alternative name of steepness degeneracy @xcite . .
in this particular example , a value of @xmath20 was used in the construction of a degenerate solution according to equation ( [ eq : sheetdegen]).,scaledwidth=48.0% ] as a concrete example , let us call the mass distribution of the left panel of fig .
[ fig : monopole ] @xmath21 , producing a deflection field @xmath22 . using equation ( [ eq : alphasymm ] )
, one can show that for a sheet of mass of precisely the critical density @xmath23 , the deflection angle simply becomes : @xmath24 if one then constructs a new mass density @xmath25 as follows : @xmath26 it is a straightforward exercise to show that for the combined deflection @xmath27 the new lens equation is : @xmath28 i.e. merely a rescaling of the original one .
this means that a rescaled version of the source plane corresponds to the same images , as is illustrated in fig .
[ fig : sheetdegenimages ] .
if the source involved is variable , it may be possible to measure time delays between the images : @xmath29 where @xmath30 and @xmath31 are the positions of two of the images . for a sheet of mass consisting of precisely the critical density ,
the projected potential is @xmath32 so that the mass sheet degeneracy transforms an initial lens potential @xmath33 into @xmath34 noting that the mass sheet degeneracy changes a source position @xmath35 to @xmath36 , a brief calculation shows that the relationship between the original time delay @xmath37 and the time delay of the degenerate version @xmath38 becomes : @xmath39 this means that time delay measurements break this mass sheet degeneracy , since a particular version of the degeneracy corresponds to a particular time delay . and @xmath40 .
the locations of the images of each source are indicated on these profiles .
the dashed line , an interpolation between the two curves , shows a total mass profile that effectively behaves as the two sheets , from the perspective of the images ( see text).,scaledwidth=49.0% ] the mass sheet degeneracy also has a particularly simple effect on the magnification factor @xmath18 . since each dimension
is scaled by a factor @xmath41 , the surface area of a small source is scaled by a factor @xmath42 . keeping the sizes of the images constant
, this means that the new magnifications of the source are given by @xmath43 .
this result can also be obtained by combining the general expression for the magnification ( [ eq : magnification ] ) and the relation between the degenerate versions of the lens equation ( [ eq : sheetdegenlenseq ] ) .
a first generalization of the mass sheet degeneracy was described in @xcite .
a brief review shall be presented here , before discussing an even more general version of this degeneracy .
it has often been claimed that having two distinct sources at different redshifts breaks the mass sheet degeneracy . from the construction above ,
the reason for this is clear .
a sheet with precisely the critical density was used , but since the critical density is redshift dependent , each source would now require the use of a sheet with a different density .
this situation is illustrated in fig .
[ fig : doublesheet ] .
source @xmath44 would require a sheet of mass which gives rise to the enclosed mass profile @xmath45 .
similarly , source @xmath46 would need a sheet with mass profile @xmath47 .
the radii at which images of each source are present are indicated on the corresponding mass profile . while it is true that a single mass sheet can no longer be used for a construction like the one in equation ( [ eq : sheetdegen ] ) , it indeed is possible to construct a mass density that for each source has the same effect as its corresponding mass sheet .
the key point is that for a circularly symmetric mass distribution , equation ( [ eq : alphasymm ] ) shows that the deflection angle is caused by the total enclosed mass within a specific radius , without any dependence on the structure of the mass profile within said radius .
this implies that if one were to construct the mass profile indicated by the dashed line in fig .
[ fig : doublesheet ] , the deflection angle at the location of the images of source @xmath44 will be the same as when caused by a mass sheet of precisely its critical density .
similarly , the images of source @xmath46 will effectively ` see ' a mass sheet with their critical density .
the mass density corresponding to this new profile can then be used to construct a degenerate solution in the same way as in equation ( [ eq : sheetdegen ] ) .
calling @xmath48 and @xmath49 the radii at which images of sources @xmath44 and @xmath46 are located , by construction this implies that : @xmath50 the effect is the same as for the original mass sheet degeneracy : the images will now correspond to sources which are rescaled by the same factor @xmath41 .
the magnifications are therefore again modified by @xmath51 . the effect on the time delay
is not as straightforward however . for a circularly symmetric lens ,
combining equations ( [ eq : lenspotential ] ) and ( [ eq : alphasymm ] ) shows that @xmath52 i.e. the lens potential , and therefore the time delays , will depend on the precise manner in which the interpolation between the two curves in fig .
[ fig : doublesheet ] is done . using the generalization of the mass sheet degeneracy that was just shown , the sources are still rescaled with the same factor .
but it is also possible to construct a degenerate solution which allows one to scale each source with a different factor . to illustrate this ,
consider the situation shown in the left panel fig .
[ fig : sheetmostgeneral ] : an elliptical mass distribution causes two sources to produce three images each . in
what follows , only source @xmath44 will be rescaled with a certain factor . in a subsequent step
the other source can be rescaled in a similar way , using a different factor . focusing on only one source and
temporarily forgetting about the other , the source can be rescaled using the classic mass sheet degeneracy from equation ( [ eq : sheetdegen ] ) .
this means that to the initial @xmath21 , a mass distribution @xmath53 was added to obtain @xmath25 : @xmath54 { \mathrm{.}}\ ] ] taking the second source back into account , it is clear that adding @xmath53 to the original mass distribution will cause the images of source @xmath44 to correspond to a scaled version of that source , but the effect on the second source is less desirable .
however , suppose it is possible to modify @xmath53 so that the deflection it causes for source @xmath44 is unaltered , and no deflection is caused at the location of the images of source @xmath46 . in that case , after adding @xmath53 , the images of source @xmath44 would correspond to a scaled version of that source , and the images of source @xmath46 would still be projected onto the same source area .
calling @xmath55 and @xmath56 the points of which respectively the images of sources @xmath44 and @xmath46 consist , @xmath57 by equations ( [ eq : lenspotential ] ) and ( [ eq : convergence ] ) implies that @xmath58 .
if one then constructs @xmath59 , in general one finds that @xmath60 starting from @xmath25 , in a next step one can then exchange the roles of the two sources and rescale source @xmath46 with a different factor .
there still remains the question of how to modify @xmath53 so that it does not produce a deflection at the location of the images of source @xmath46 .
this can be done by using a large number of the monopole basis functions we met earlier . if none of these overlap with the images of source @xmath44 , the deflection there will be unaltered . to make sure that the deflection angle at the location of the images of source @xmath46 vanishes
, appropriate weights of these basis functions must be sought .
an example of this approach can be seen in the center panel of fig .
[ fig : sheetmostgeneral ] . in this example , the weights were determined by a genetic algorithm , an optimization strategy inspired by natural selection .
a similar procedure is used as in @xcite : a trial solution consists of a set of weights for the monopole basis functions . starting from randomly initialized trial solutions , the algorithm then ` breeds ' increasingly better trial solutions by combining existing solutions and introducing mutations .
a key step in this procedure is the application of selection pressure : better solutions should produce more offspring . in this case ,
a trial solution is deemed better than another simply when its deflection field differs less from the desired one at the locations of the images of source @xmath46 . since , when generating a degenerate solution in this fashion , one still uses a rescaled version of the original mass distribution , this procedure still produces a mass map which appears rescaled , as can be seen in the center panel of fig .
[ fig : sheetmostgeneral ] .
another approach which does not cause such an apparent rescaling , can be used as well . from the discussion above
, it is clear that the final version of the mass distribution @xmath53 should produce a specific value of the deflection angle @xmath61 at the location of the images of source @xmath44 . on the other hand , at the location of the images of source @xmath46 , no deflection should be produced .
we therefore know the deflection angle at specific locations , and at other locations the deflection angle can take a large number of values , the only real constraint being that no additional images should be produced .
interpolating a deflection field in such a way that @xmath62 , or equivalently that the deflection originates from a scalar lensing potential , is precisely what is done in the lensperfect lens inversion procedure @xcite . using a similar interpolation procedure to determine @xmath61 and therefore @xmath53 ,
yields the result shown in the right panel of fig .
[ fig : sheetmostgeneral ] . while more difficult to ensure that the overall mass distribution is positive , in this example it produces an interesting result .
comparing the resulting mass distribution with the original one in the left panel , one sees that the differences are rather subtle . and
since no rescaling of the mass density itself seems to be involved , one would not suspect a mass sheet - like degeneracy to be at work here .
note that in principle , a similar construction is possible when only using point sources . when the deflection angles are only conserved at very specific point locations and not over a more extended area , the gradients of the deflection field will change and therefore also the surface density and the magnification factors .
in particular , the relation in equation ( [ eq : almostgeneralsheetdegen ] ) will no longer hold . however , since this no longer corresponds to a rescaling of the sources , it can hardly be considered as caused by the mass sheet degeneracy .
to further illustrate the significant practical problems posed by these degeneracies , we shall use a simulated lensing scenario .
the left panel of fig .
[ fig : experimentsetup ] shows the true mass map used in the simulation , an elliptical generalization of a nfw mass distribution @xcite , with concentration parameter @xmath63 and scale parameter @xmath64 arcseconds .
the lens was placed at a redshift of @xmath65 in a cosmology with @xmath66 km s@xmath67 mpc@xmath67 , @xmath68 and @xmath69 .
the surface density is expressed in terms of the critical density for a source at @xmath70 .
eight sources were placed at different redshifts .
these generate 26 images which can be seen in the contour map , and also in the right panel of the same figure .
the redshifts of the sources lie between @xmath71 and @xmath72 .
the center panel of fig .
[ fig : experimentsetup ] shows the critical lines and caustics for a source at @xmath70
. we shall be interested in comparing the scale of the caustics here to that of reconstructions , as rescaling the source plane is the hallmark of the mass sheet degeneracy . first , a reconstruction is done using the algorithm presented in @xcite . in this method ,
plummer @xcite basis functions are arranged according to an adaptive grid , and the weights of these basis functions are determined using a genetic algorithm .
the left panel of fig .
[ fig : invnosheet ] shows the resulting mass distribution , an average of 20 individual reconstructions .
overall , it has a similar appearance as the input mass map shown in fig . [
fig : experimentsetup ] , and its critical lines ( center panel ) do bear resemblance to those of the true lens .
the resulting mass distribution , however , is considerably steeper than the one used in the simulation , which can also clearly be seen in the right panel of fig .
[ fig : invnosheet ] . in this image
we show the densities measured at the center of each image . for the sake of visualization , instead of using the two dimensional coordinate of each image center point , the distance to the nfw center is used on the @xmath73-axis . comparing the two profiles on that figure , one
is immediately reminded of the mass sheet degeneracy , i.e. the profiles seem to differ mostly by a rescaling and the addition of a specific offset .
similarly , when looking at the caustics in the center panel of fig .
[ fig : invnosheet ] , they seem to be an expanded version of the ones of the input lens , again a feature that hints at the presence of the mass sheet degeneracy . due to the inherent uncertainty in between the images caused by the monopole degeneracy
, the most sensible thing to do in trying to retrieve the nfw parameters , is to perform a fit based on the densities precisely at the locations of the images .
when looking at the steepness of the reconstructed mass map , it will come as no surprise that extracting the concentration and scale parameters from the reconstructed densities will not yield accurate results . in this particular case ,
the estimates are @xmath74 and @xmath75 arcseconds respectively ( 68% confidence level ) .
using plummer basis functions it is of course rather difficult to mimic the effect of a sheet of mass .
given the fact that within the images a considerable mass sheet - like component seems to be present in the input lens , the fact that the reconstruction is rather different is not that surprising . in a second reconstruction
not only plummer basis functions were used , but a single mass sheet basis function was present as well , as is also explained in @xcite .
the resulting mass distribution , again an average of 20 individual solutions , can be seen in the left panel of fig .
[ fig : invsheet ] .
this mass distribution is clearly a lot more similar to the input mass map , also obvious from the right panel of the same figure .
as can be expected from the similarity between the true and reconstructed lenses , the caustics shown in the center panel no longer show a clear difference in scale . when performing a fit in a similar fashion as the previous reconstruction , one finds nfw parameters of @xmath76 and @xmath77 arcseconds , corresponding better to the true values .
unfortunately , in this example there is no real reason to prefer one reconstruction over the other , meaning that apart from the overall shape of the true mass distribution , little would be revealed by performing these inversions .
furthermore , in both cases fitting the nfw shape yields an overestimate of the concentration parameter and an underestimate of the scale radius , although the effect is far more serious in the first inversion . earlier in this article it was mentioned that time delay information can help not only in breaking the most basic mass sheet degeneracy ; also for the generalized versions time delays depend on the precise construction method of the degenerate mass map . in @xcite
a method was described with which time delay information can be added as a constraint , now using a so - called multi - objective genetic algorithm . to explore the value of such time delays ,
time delay information of a single system of images was added to the inversion procedure .
the three images used for this added constraint are indicated in the right panel of fig .
[ fig : experimentsetup ] and correspond to a lensed source at redshift 2.75 .
first , the inversion was run using plummer basis functions alone , i.e. no mass sheet basis function was used .
the resulting mass distribution can be seen in the left panel of fig .
[ fig : invnosheettd ] . concerning the input to the inversion procedure , the only difference with the results obtained in fig . [
fig : invnosheet ] is the addition of time delay information .
the effect of this information is impressive however : especially when comparing the right panels of these two figures it is clear that the time delay constraint seems to pull the densities at all image locations towards the true densities .
the caustics in the center panel confirm the mass sheet - breaking effect : the overall scale of the source plane seems to match the true one more closely .
note that this is solely the effect of time delays for a single image system .
when determining the nfw parameters as before , one now finds @xmath78 and @xmath79 arcseconds . this result still does not reflect the true parameters , but is far less deviant than the estimates based on fig .
[ fig : invnosheet ] . as a final inversion ,
both the time delay information and the mass sheet basis function was used , i.e. the inversion is similar to the one reflected by fig .
[ fig : invsheet ] , but now including time delay information about one set of images . the resulting mass map is depicted in the left panel of fig . [
fig : invsheettd ] ; the densities at the image locations can be seen in the right panel of the same figure .
the effect of the added time delay information is not as impressive as in the previous example , but still it leads to an improvement of the estimated densities at most image points .
only the density at the point closest to the nfw center seems to be underestimated systematically .
this single outlier does not seem to have a profound adverse effect on the estimation of nfw parameters however , which now are @xmath80 and @xmath81 arcseconds , very close to the true values .
the examples shown in this article make it clear that these lensing degeneracies are closely related to substructure in the mass density .
the monopole degeneracy is the worst , as it has no effect on any property of the images : the same sources predict the same images , with identical magnifications and identical time delays .
in essence it means that we do not have a firm handle on the mass distribution in between the images , the only real constraint being the absence of unobserved images .
this illustrates both the importance of gravitational lens systems with a multitude of images and when necessary , of the use of the null space , i.e. the locations where no images are observed . in its most general version
, the mass sheet degeneracy does not require an actual sheet of mass , nor does it necessarily modify the slope of the mass distribution , as the alternative name of steepness degeneracy might suggest . comparing equations ( [ eq : sheetdegen ] ) , ( [ eq : doublesheetdegen ] ) and ( [ eq : almostgeneralsheetdegen ] ) , it becomes clear that when scaling source @xmath44 with factor @xmath82 and source @xmath46 with factor @xmath83 , the effect on the mass density is the following : @xmath84 where @xmath55 and @xmath56 are again the locations of the images of source @xmath44 and @xmath46 respectively .
of course , this can be generalized to any number of sources and images .
apart from rescaling each source in the system , equation ( [ eq : generalsheetdegen ] ) illustrates that this degeneracy changes the density precisely at the locations of the images , i.e. it necessarily introduces substructure .
the effect on the time delays is in general not straightforward implying that time delay information may be very valuable in helping to break this degeneracy .
the simple effect on the magnification is not easy to use since one most often does not have information about the size of the source .
efforts to use the magnification information to avoid the mass sheet degeneracy have been undertaken however , e.g. in @xcite .
the fact that both monopole and mass sheet degeneracies are at least in principle
possible with any amount of sources and images , combined with the fact that they are both linked to substructure in the mass distribution , means that in reality it is the prior information on the degree of smoothness of the mass distribution that effectively breaks these degeneracies .
the degree of smoothness can depend implicitly on the method used , e.g. in parametric methods like the lenstool method @xcite , the method of @xcite which is largely based on the distribution of the light , or the method of the authors in which overlapping smooth basis functions are used ( e.g. @xcite ) .
other inversion procedures use an explicit regularization scheme or an explicit bayesian prior , e.g. in the pixelens method ( @xcite , @xcite ) .
the dependence on explicit or implicit assumptions about the smoothness of the solution of course means that one must be careful when interpreting obtained inversion results , not only in lensing systems with few sources , but also when a larger amount of sources are available , as indicated by the experiment .
this same experiment also highlights the importance of time delay information , which easily asserts a global effect on the reconstruction .
it is unfortunate that in many cases of interest these time delays may simply be too long to measure in practice .
for the inversions performed in this article we used the infrastructure of the vsc - flemish supercomputer center , funded by the hercules foundation and the flemish government - department ewi . | the inversion of gravitational lens systems is hindered by the fact that multiple mass distributions are often equally compatible with the observed properties of the images . besides using clear examples to illustrate the effect of the so - called monopole and mass sheet degeneracies
, this article introduces the most general form of said mass sheet degeneracy .
while the well known version of this degeneracy rescales a single source plane , this generalization allows any number of sources to be rescaled .
furthermore , it shows how it is possible to rescale each of those sources with a different scale factor . apart from illustrating that the mass sheet degeneracy is not broken by the presence of multiple sources at different redshifts
, it will become apparent that the newly constructed mass distribution necessarily alters the existing mass density precisely at the locations of the images in the lens system , and that this change in mass density is linked to the factors with which the sources were rescaled . combined with the fact that the monopole degeneracy introduces a large amount of uncertainty about the density in between the images ,
this means that both degeneracies are in fact closely related to substructure in the mass distribution .
an example simulated lensing situation based on an elliptical version of a navarro - frenk - white profile explicitly shows that such degeneracies are not easily broken by observational constraints , even when multiple sources are present .
instead , the fact that each lens inversion method makes certain assumptions , implicit or explicit , about the smoothness of the mass distribution means that in practice the degeneracies are broken in an artificial manner rather than by observed properties of the lens system .
[ firstpage ] gravitational lensing dark matter |
eta carinae s photometric record is unparalleled among well - studied objects , especially since it has been near or exceeded the classical eddington limit during the past two or three centuries . from 1700 to 1800
it gradually brightened from 4th to 2nd magnitude , and then experienced its famous great eruption or `` supernova impostor event '' beginning about 1837 . for twenty years it was one of the brightest stars in the sky , rapidly fluctuating between magnitudes 1.5 and 0.0 , briefly attaining @xmath1 .
after 1858 it faded below 7th magnitude , presumably enshrouded in the nascent homunculus nebula .
subsequent behavior , however , has been more complex than one might have expected .
a mysterious secondary eruption occurred in 18871900 ; then the apparent brightness leveled off around @xmath2 for about 40 years , followed by a rapid increase in 193853 ; after that it brightened at a fairly constant rate for another 40-year interval , and most recently the rate accelerated in the 1990 s .
some , but not all , of the secular brightening can be attributed to decreasing obscuration as the homunculus nebula expands . however , in truth this is more complex than it appears .
the star s physical structure has been changing in a decidedly non - trivial way which is , at best , only dimly understood . for historical and observational details
see @xcite .
spectroscopic changes have occurred along with the brightness variations .
the 5.5-year spectroscopic / photometric cycle @xcite is not apparent in data obtained before the 1940 s @xcite .
brief `` spectroscopic events '' marking the cycle are most likely mass - ejection or wind - disturbance episodes , probably regulated by a companion star @xcite . at visual wavelengths ,
the associated ephemeral brightness changes represent mainly emission lines in the stellar wind , while the longer - term secular brightening trend involves the continuum @xcite .
@xcite , @xcite , and @xcite have speculated that the four obvious disruptions in the photometric record c. 1843 , 1893 , 1948 , and 2000 might indicate a quasi - periodicity of the order of 50 years . in any case
the star has not yet recovered from its great eruption seen 160 years ago .
the hubble space telescope ( hst ) treasury program for @xmath0 car was planned specifically to study the 2003.5 spectroscopic event .
we employed the space telescope imaging spectrograph ( stis ) and advanced camera for surveys ( acs ) , following earlier stis observations that began in 1998 @xcite .
fortuitously , the stis data almost coincide with a rapid secular brightening which began shortly before 1998 ( see section 4 below ) .
those and the acs images are of unique photometric value for at least two reasons : 1 . at visual wavelengths , normal ground - based observations have been dominated by the surrounding homunculus ejecta - nebula , which , until recently , appeared much brighter than the central star and which has structure at all size scales from 0.1 to 8 arcseconds .
_ so far , only the hst has provided well - defined measurements of just the central star . _
the homunculus is primarily a reflection nebula , but the homunculus / star brightness ratio has changed substantially . during 1998 - 99 , for instance , the star nearly tripled in apparent brightness while ground - based observations showed only about a 0.3-magnitude brightening of homunculus plus star @xcite .
this rather mysterious development is known from hst / stis and hst / acs data .
numerous strong emission lines perturb the results for standard photometric systems .
h@xmath3 and h@xmath4 emission , for example , have equivalent widths of about 800 and 180 respectively in spectra of @xmath0 car .
broad - band @xmath5 , @xmath6 , @xmath7 , and @xmath8 magnitudes , and most medium - band systems as well , are therefore poorly defined for this object .
photometry around 5500 , e.g. broad - band @xmath9 , is relatively free of strong emission lines , but transformations from instrumental magnitudes to a standard system require the other filters @xcite .
this difficulty is somewhat lessened for hst observations restricted to the central star , whose spectrum has fewer emission lines than the bright ejecta ; and some of the hst / acs filters are fairly well - adapted to the case . at any rate
the stis and acs data appear to be stable and internally consistent . detector and filter systems used in most ground - based work , on the other hand , require fluctuating instrumental and atmospheric corrections , and do not give any major advantage for this object .
in this paper we present the complete set of photometric data gathered for the @xmath0 car hst treasury project .
the central star has brightened , especially in the uv , since the hst results described by @xcite .
we report three types of later measurements : 1 .
the star s brightness in acquisition images made with stis before that instrument s failure in early 2004 .
these images represent a broad , non - standard wavelength range from 6500 to 9000 .
2 . the star s brightness in acs / hrc images made in four filters ( f220w , f250w , f330w , & f550 m ) from october 2002 to the present .
3 . synthetic photometry in flux - calibrated stis spectra . after presenting the data below
, we briefly discuss the observed trends and what they bode for the near future of @xmath0 car .
our principal reason for reporting these data now is that the treasury program observations have been completed ; future hst / acs observations are possible but not assured .
these last treasury program observations are essential for demonstrating the secular trend in brightness ( see section [ trend ] ) , that these brightness changes suggest a fundamental change in the state of the star ( see section [ discussion ] ) and what the near future may hold for @xmath0 carinae ( [ predictions ] ) .
each set of stis observations included a pair of acquisition images , which are 100 x 100 pixel sub - frames ( @xmath10 ) centered on the middle row and column of the ccd @xcite .
the stis acquisition images were taken with a neutral density filter ( f25nd3 ) which , combined with the ccd response , covered the wavelength range 200011000 . since the star s apparent color is moderately red ,
these images were dominated by fluxes at wavelengths 65009000 .
[ filterfig ] shows that although most of the measured flux comes from the continuum , several prominent emission features including h@xmath3 contributed to the measured brightness .
we measure the star s brightness with in radius @xmath7 in the following manner : if @xmath11 denotes the flux level in an image where the star was centered at @xmath12 , @xmath13 , we integrate the product of @xmath14 , where @xmath15 is a radially symmetric weighting function of the form @xmath16,@xmath17 . in effect
@xmath15 is a `` parabolic virtual field aperture . '' for the stis acquisition images we chose @xmath18 ( 6 pixels ) . a detailed discussion of these reduction procedures , including a non - standard bias level correction which we applied , is given in @xcite .
initially we expected to add a few more points of acquisition photometry after 2004.5 , but the failure of the stis in august 2004 curtailed our observing plans .
thus we have only two additional stis data points to report ( table [ acqdata ] ) beyond those given in @xcite .
they are useful , however , concerning the end of the post - event recovery ( see sections [ trend ] and [ discussion ] below ) .
hst acs / hrc observations of @xmath0 car were obtained for the treasury project beginning in october 2002 .
we have also examined publicly available data from hst proposal 9721 and hst proposal 10844 .
the bias - corrected , dark - subtracted , and flat - fielded data were obtained from the space telescope science institute via the multi - mission archive ( mast ) @xcite .
treasury program acs / hrc images were taken in four filters that cover near - uv to visual wavelengths ( fig .
[ filterfig ] ) : * hrc / f220w and hrc/250w : these near - uv filters sample the `` forest '' @xcite , whose opacity increases dramatically during a spectroscopic event @xcite .
* hrc / f330w : this filter includes the balmer continuum in emission , supplemented by various emission lines .
it also attained the best spatial resolution among the observations reported here .
* hrc / f550 m : with a medium - width ( not broad ) bandpass , this filter samples the visual - wavelength continuum flux with only minor contamination by emission features .
the brightness of the central star was measured using the same 0.3@xmath19 ( @xmath20 10 acs / hrc pixels ) weighted virtual aperture used for the stis acquisition images .
ccd flux values were converted to the stmag system @xcite using the keywords provided in the mast archive s fits headers .
an aperture correction la @xcite , calculated from observations of the white dwarf gd 71 in each of the filters ( table [ acscaltab ] ) , was applied to the measurements ( table [ acsdata ] ) .
acs fluxes and magnitudes measured prior to mjd 52958 ( 2003.87 ) can be found in our first paper @xcite . _
caveat : _ we did not apply the aperture corrections to the magnitudes in that paper , but we have done so in the plots shown here . the f220w and f250w filters have known red leaks @xcite that can affect photometry of red sources .
we convolved extracted stis spectra ( see sec .
[ stisphot ] ) with the response function for those filters and the acs / hrc . in the case of the central star of @xmath0 car ,
the flux redward of 4000 in the f220w and f250w filters contributed only about 0.25% and 0.06% respectively , insignificant compared to other sources of error .
originally the acs / hrc images were meant to supplement the stis spectra . after the untimely demise of the stis , however , the acs / hrc became the most suitable mode for observing @xmath0 car with hst .
this presented a problem of continuous monitoring in the same band passes over the entire program , since there were no acs / hrc images prior to 2002.78 while the stis data ended at 2004.18 .
acs photometry can be synthesized from the flux calibrated stis data , since nearly every grating tilt was observed during each stis visit ( table [ stisspectra ] ) .
the spectra were extracted with a weighted parabolic cross dispersion profile similar to the virtual aperture used to measure the acs / hrc images , convolved with the published filter functions ( fig [ filterfig ] ) , and integrated . because stmag is computed from flux density , the integrated fluxes were divided by the effective band passes of each filter ( see table [ acscaltab ] ) .
the effective aperture for the extracted stis spectral data is not a rotated parabola but a parabolic cylinder having the width of the slit ( 0.1 ) . to correct for the difference in aperture as well as the difference in instrumental psf , slit throughput , and the extraction height , we converted the stis spectral fluxes to the acs / hrc flux scale using suitable correction factors ( table [ acscaltab ] ) .
those factors were computed by comparing the results from acs / hrc images and photometry synthesized from stis data obtained at time mjd 52683 .
the resulting synthetic acs / hrc photometry is given in table [ fakeacs ] .
plots of these various data will be discussed in sections [ trend ] and [ discussion ] below .
since hst and ground - based photometry have shown different rates of change for different - sized areas , it is useful to view the spatial distribution of the brightness . for this purpose
we have used three acs images made with filter f550 m at @xmath21 2004.93 , 2005.53 , and 2005.85 . together
these give a reasonably valid picture of the average visual - wavelength appearance during 2005 .
[ brt - distr-1 ] shows the fraction of apparent brightness within projected radius @xmath7 measured from the central star .
the solid curve represents the hst data , while a companion dashed curve incorporates gaussian blurring with fwhm = 0.8 , simulating ground - based photometry with fairly good atmospheric conditions .
half the total light originates within @xmath22 which is very different from @xmath0 car s appearance a few decades ago @xcite .
[ brt - distr-1 ] also shows a curve based on photographs that gaviola obtained in 1944 @xcite , with a magnified spatial scale to compensate for subsequent expansion ; even allowing for mediocre `` seeing '' and other uncertainties , the degree of central condensation was obviously less then . before 1980
the central star accounted for less than 10% of the total apparent brightness ; now its fraction has grown to about 40% and continues to increase .
[ brt - distr-2 ] is a map of the surface brightness , based on the hst / acs f550 m data mentioned above . in order to produce simple well - defined isophotes ,
we have gaussian - blurred the image using fwhm 0.5 . apart from the central star and nearby compact ejecta
, most of the light comes from a comma- or crescent - shaped region about 5 across , marked by the 80% isophote in the figure .
presumably the high intensity in this area results from strong forward scattering by dust grains in that part of the southeast homunculus lobe . surface brightnesses in the outer lobe regions , on the other hand , are fainter than the 50% contour by factors typically between 100 and 200 . about half of the projected area of the homunculus provides only 5% of the total visual - wavelength brightness .
fig . [ plot1 ] shows the main hst photometric data on the central star ; the most significant result is a secular brightening trend superimposed on the 5.5-year pattern of spectroscopic events .
the latest observations are essential in this regard , because the data reported earlier by @xcite ended before the star had emerged from the 2003.5 event and we could not be sure of the long term trend . from 1999 to
2006 the average trend was about 0.15 magnitude per year at visual wavelengths and 0.19 mag yr@xmath23 at 2200 much faster than the rate of @xmath20 0.025 mag yr@xmath23 recorded for the homunculus plus star from 1955 to 1995 @xcite .
the weigelt condensations northwest of the central star @xcite have _ not _ brightened rapidly . located in the equatorial plane only about 800 au from the star car s distance
is 2300 pc @xcite .
] , their light is intrinsic emission with some reflection @xcite .
[ knotd ] shows the brightness of `` weigelt d , '' measured in the same way as the star but centering the virtual aperture at offset location @xmath24 , position angle 336@xmath25 .
the absence of a strong secular trend is significant in the following way .
extrapolating the recent trend of the star / ejecta brightness ratio back to the mid-1980 s , one would expect that the star should have been fainter than each of the weigelt blobs at that time .
but this is contradicted by early speckle observations @xcite ; _ therefore the star can not have brightened at the present - day rate through the entire 20-year interval .
_ moreover , the earliest hst / fos spectroscopy in 1991 @xcite and the ultraviolet spectra of the star plus inner ejecta obtained with the international ultraviolet explorer ( iue ) from 1979 to 1990 show absolute fluxes which , though uncertain , appear comparable to both the speckle observations and the 1998 stis results .
these facts imply that the central star s brightening rate was relatively modest from 1980 until sometime in the 1990 s .
we suspect that the present - day rate began in 199497 , when ground - based photometry showed unusual behavior ( see , e.g. , fig . 2 in @xcite ) .
the last f330w and f550 m observations in fig .
[ plot1 ] confirm the sudden 0.2 magnitude increase observed at la plata in late 2005 @xcite .
the observed brightening of @xmath0 car is not easy to explain . it can not signify a major increase in the star s luminosity , because that would exceed the eddington limit , producing a giant eruption .
it can not be a standard lbv - like eruption ; in that case the energy distribution should have shifted to longer wavelengths , the balmer emission lines should have decreased , and the spectrum should have begun to resemble an a- or f - type supergiant @xcite . in fact , qualitatively the star s spectrum has changed little in the past decade , and it has become bluer , not redder .
car has long been known to have an abnormally small reddening / extinction ratio , see @xcite , @xcite , and refs .
cited therein . ]
the most obvious remaining explanation involves a change in the circumstellar extinction , which , in turn , probably requires a subtle change in the stellar wind .
mere `` clearing of the dust ''
i.e. , motion of a localized concentration of dusty ejecta can not occur fast enough @xcite
. therefore one must consider either destruction of dust grains , or a decrease in the formation of new dust , or both ; and , if these account for the observations , why should they happen now ?
the hypothetical decreasing extinction probably occurs within 2000 au ( @xmath20 1 ) of the star , and preferably closer , because : 1 . in various observations between 1980 and 1995
, the star did not appear as bright as expected relative to the weigelt blobs ; the discrepancy was a factor of the order of 10 , based on simple theoretical arguments @xcite . evidently , then , our line of sight to the star had substantially larger extinction at visual wavelengths , even though its projected separation from the blobs was less than 0.3 .
the required extra extinction was of the order of 3 magnitudes .
since then the star has brightened far more than the weigelt objects have ; therefore , if this involves localized extinction , its size scale must be a fraction of an arcsec , only a few hundred au .
no known process seems likely to destroy dust more than 2000 au from the star in a timescale of only a few years .
ground - based photometry and hst images have shown only a modest , fraction - of - a - magnitude increase in the brightness of the large - scale homunculus lobes during the past decade @xcite .
dust grains should condense in @xmath0 car s wind at a distance of 200600 au , 2 to 10 years after the material has been ejected .
car , we suppose that appreciable grain condensation begins in the outward flowing material at the location where the equilibrium grain temperature is around 1000 k. this is a fairly conventional assumption and the precise choice of temperature has little effect on our reasoning .
the quoted time - after - ejection assumes typical ejecta speeds of 200700 km s@xmath23 . ] since newly - formed dust moves outward in a timescale of several years , the circumstellar extinction seen at any time depends partly on the current dust formation rate .
this , in turn , depends on local wind density , radiation density , etc . , and newly formed hot grains ( @xmath26 k ) are susceptible to destruction .
the dust column density can thus be sensitive to small changes in the stellar parameters .
moreover , the wind is latitude - dependent and our line of sight is close to the critical latitude where wind parameters can vary rapidly @xcite .
all these factors appear suitable for the proposed explanation . on the other hand ,
near - infrared observations imply that extinction within @xmath27 au has been quite small along most paths outward from the star . in fig . 3 of @xcite , for instance , the 26 @xmath28 m flux indicates the high end of the dust temperature distribution . modeling this in a conventional way
, we find that less than 5% of the total luminosity was absorbed and re - emitted by inner dust with @xmath29 k during the years 1973 to 1990 when those observations were made . at wavelengths around 4 @xmath30 m . assuming a typical emission efficiency dependence @xmath31 , the observed spectral slope can be explained by a grain temperature distribution @xmath32 .
the result noted in the text is obtained by normalizing this to match the observed flux around 4 or 5 @xmath28 m and then integrating the total emitted flux at all wavelengths due to grains above 500 k. ] therefore , _ our line of sight must be abnormal in order to have a large amount of extinction near the star . _ in principle one might view this as an argument against our proposed scenario , but no plausible alternative has been suggested to explain the apparent faintness of the central star before 1999 and its ratio to the weigelt blobs @xcite , the spatial distribution of dust is probably quite inhomogeneous near the star .
the homunculus lobes have a conspicuously `` granular '' appearance ; the equatorial ejecta are clumpy , including the weigelt knots ; and stars near and above the eddington limit tend to produce clumpy outflows @xcite .
consequently the radiative transfer problem includes macroscopic effects which have not yet been modeled .
if the grain albedo is sufficient , light may escape mainly by scattering through interstices between condensations . in that case , high - extinction lines of sight may be fairly common in the inner region even though most of the light escapes along other paths , not necessarily radial .
incidentally , the near - infrared photometric trends reported by @xcite are not straightforward to interpret . the fairly - constant 3.5 @xmath28 m flux , for instance ,
represents a complicated mixture of dust formation parameters and does not necessarily indicate a constant amount of dust ; see comments by @xcite .
if the observed brightening represents a decrease in circumstellar extinction , the likeliest reason for this to occur is through some change in the star
no one has yet proposed a suitable alternative .
the most relevant stellar parameters are the radius , current luminosity , and surface rotation rate , which together determine the wind s velocity , density , and latitude structure .
all of these may still be changing today , 160 years after the great eruption ; thermal and rotational equilibrium in particular are likely to be poor assumptions for the star s internal structure @xcite . as a working hypothesis to explain @xmath0 car s photometric and spectroscopic record in the past 100 years ,
let us tentatively suppose that _ the mass - loss rate is gradually decreasing , _ while the surface rotation rate may be increasing .
historical considerations include : 1 .
high - excitation emission , now observed at most times , was consistently absent before 1920 @xcite and probably before 1940 @xcite .
if a hot companion star is present as most authors suppose , then the most obvious way to hide or suppress its helium ionization is to immerse the entire system in an extremely dense wind
i.e. , the primary star s mass - loss rate must have been larger then .
this idea is far from straightforward @xcite , but so far as we know it is the only qualitative explanation yet proposed .
informally , based on zanstra - style arguments ( i.e. , assessing the volume emission measure @xmath33 needed to absorb all the photons above 25 ev ) , we estimate that a rate of the order of 10 times the present value , i.e. @xmath34 @xmath35 y@xmath23 , would have been required early in the twentieth century in order to suppress the helium recombination emission .
twenty years ago the amount of fresh dust , indicated by the near - infrared flux , appeared consistent with a mass - loss rate somewhat above @xmath36 @xmath37 yr@xmath23 @xcite .
this absorbed only a small fraction of the luminosity ( section 5.1 above ) , but the substantially higher mass - loss rate suspected for earlier times would have produced enough hot inner dust to absorb a non - negligible fraction .
3 . the brightness observed between 1900 and 1940 is rather mysterious .
judging from its mass and present - day optical thickness , around 1920 the homunculus ( then only half as large as it is today ) should have had at least 5 magnitudes of visual - wavelength extinction ; in a simple model the object should have been fainter than 10th magnitude instead of @xmath2 as was observed .
no doubt the inhomogeneities mentioned earlier played a role , but no model has been calculated .
moreover , why did the brightness remain fairly constant even though the homunculus expanded by about 70% in 19001940 ?
this interesting problem has received practically no theoretical attention .
emission first appeared , and @xmath0 car s brightness suddenly increased , between 1938 and 1953 as we mentioned in section 1 .
this might conceivably be explained by a decrease in the wind density ; but @xcite and @xcite have conjectured that 19401950 may have been the time when rotation became fast enough to produce latitude structure in the wind .
if so , a higher - excitation , lower - density zone then developed at low latitudes @xcite .
the above points inspire two hypotheses that may explain the rapid brightening trend shown in fig .
[ plot1 ] .
first , if the mass - loss rate has been decreasing , this tends to reduce the column density of recently - formed dust along our line of sight . meanwhile ( or alternatively ) , perhaps the wind s latitude structure is continuing to evolve so that its dense zone is now moving out of the line of sight .
hst data suggest that our line of sight has been fairly close to the critical boundary latitude separating the two phases @xcite .
a small increase in surface rotation rate , or some other parameter change , might conceivably move the dense zone to higher latitudes , decreasing the amount of dust that forms along our line of sight .
this idea is appealing because it suggests a way in which the effective extinction may be very sensitive to the stellar parameters .
this problem obviously requires detailed models far beyond the scope of this paper , combining the star s changing structure , its wind , dust formation , and possibly dust destruction . figs .
[ plot1 ] and [ plot2 ] reveal no major surprises about the 2003.5 spectroscopic event , but several comments are worthwhile .
first , the sharp drop in uv brightness ( filters f220w and f250w ) is qualitatively understood and does not involve circumstellar dust . during both the 1998 and the 2003 events , stis data showed very heavy ultraviolet blanketing by ionized metal lines ; indeed the star became quite dark at some wavelengths between 2000 and 3000 @xcite .
we further note that just before the spectroscopic event , a slight increase occurred at wavelengths below 4000 ( filters f220w , f250w , f330w ) , but not at visual and far - red wavelengths ( f550 m and f25nd3 ) .
ground - based visual - wavelength and near - ir photometry showed a qualitatively similar effect @xcite .
the brightening is particularly prominent in j , h , and k which are dominated by free - free emission @xcite .
the acs f550 m and stis f25nd3 data primarily measure the continuum brightness , while the other hst filters are heavily influenced by strong emission or absorption lines . at about the same time emission in
the central star also went through a similar increase in brightness @xcite .
the minor pre - event brightening thus appears to represent an increase in some emission features implying a temporary increase in ionizing flux . the primary star may provide additional uv photons or the hypothetical hot companion star may excite the primary wind more than usual at that time ( just before periastron ) , but no quantitative model has been attempted . figs .
[ plot1 ] and [ plot2 ] contain interesting hints about the timescale for the star s post - event recovery .
four months after the 2003.5 event , for instance , the 210 kev x - ray flux had increased almost to a normal level @xcite . the hst / acs f220w and f250w brightnesses , however , were still quite low at that time , and they required about eight months to recover .
this timescale must be explained in any valid model for the spectroscopic events .
@xcite noted serious differences between stis spectra of the 1998.0 and 2003.5 events , and interpreted them as evidence for a rapid secular physical change in @xmath0 car .
@xcite had earlier found that emission became progressively weaker after each of the last few spectroscopic events .
these clues are obviously pertinent to our comments in section [ stellarparams ] above
. fluctuations _ between _ spectroscopic events have received little attention in the past .
for instance , fig .
[ plot1 ] shows a brief 0.2-magnitude brightening at 2001.3 ; measured by the stis in both imaging and spectroscopic mode .
it was correlated with the behavior of a strange unidentified emission line near 6307 , and with other subtle changes in the spectrum @xcite .
this is interesting because mid - cycle events have not been predicted in any of the competing scenarios for the 5.5-year cycle .
perhaps the effects seen in 2001 indicate the level of basic , lbv - like fluctuations in @xmath0 car .
the appearance of @xmath0 car and the homunculus nebula has changed dramatically .
twenty years ago the entire object could have been described as `` a bright , compact nebula having an indistinct eighth - magnitude central core ; '' but a few years in the future , if recent trends continue , it will be seen instead as `` a fifth- or even fourth - magnitude star accompanied by some visible nebulosity . ''
meanwhile the color is gradually becoming bluer .
this overall development has long been expected @xcite , but now appears to be occurring 20 years ahead of schedule .
if it signals an irregularity in the star s recovery from the great eruption , then this may be a highly unusual clue to the highly abnormal internal structure .
unsteady diffusion of either the thermal or the rotational parameters would be significant for stellar astrophysics in general .
there are several practical implications for future observations of this object .
valid ground - based spectroscopy of the star ( strictly speaking its wind ) is becoming feasible for the first time , as its increased brightness overwhelms the emission - line contamination by inner ejecta .
unfortunately this implies that the inner ejecta particularly the mysterious weigelt knots are becoming difficult to observe .
in fact , since the hst / stis is no longer available , they are now practically impossible to observe .
when some new high - spatial - resolution spectrograph becomes available in the future , the inner ejecta will probably be much fainter than the star .
the expected future of the larger - scale homunculus nebula is also interesting . at present
it is essentially a reflection nebula .
however , based on the presence of high - excitation emission lines such as [ ] close to the star , the system almost certainly contains a source of hydrogen - ionizing photons with energies above 13.6 , and helium - ionizing photons above 25 ev .
( see , e.g. , @xcite ; most recent authors assume that this source is a hot companion star . )
eventually , when circumstellar extinction has decreased sufficiently due to expansion and other effects , the uv source will begin to photoionize the homunculus .
this is especially true if the primary stellar wind is weakening as we conjectured above .
first the inner `` little homunculus '' will become a bright compact region , and then the bipolar homunculus lobes .
the time when that will occur is not obvious , but it may be within the next few decades if current trends continue .
this research was conducted as part of the hst treasury project on @xmath0 carinae via grant no .
go-9973 from the space telescope science institute .
we are grateful to t.r .
gull and beth perriello for assisting with the hst observing plans .
we also especially thank roberta humphreys , j.t . olds , and matt gray at the university of minnesota for discussions and helping with non - routine steps in the data preparation and analysis .
altamore , a. , baratta , g. b. , cassatella , a. , rossi , l. , & viotti , r. 1986 , new insights in astrophysics : eight years of uv astronomy with iue , 303 cassatella , a. , giangrande , a. , & viotti , r. 1979 , , 71 , l9 clampin , m. , hartig , g. , baum , s. , kraemer , s. kinney , e. kutina , r. , pitts , r. , & balzano , v. 1996 , stis instrument science report 96 - 030 , ( baltimore : stsci ) corcoran , m. f. 2005 , , 129 , 2018 cox , p. , mezger , p. g. , sievers , a. , najarro , f. , bronfman , l. , kreysa , e. , & haslam , g. 1995 , , 297 , 168 damineli , a. 1996 , , 460 , l49 damineli , a. , stahl , o. , wolf , b. , kaufer , a. , & jablonski , f. j. 1999 , asp conf .
ser . 179 : eta carinae at the millennium , 179 , 221 davidson , k. 1987 , assl vol . 136 : instabilities in luminous early type stars , 127 davidson , k. 1999 , asp conf .
ser . 179 : eta carinae at the millennium , 179 , 304 davidson , k. 2004 , stsci newsletter , spring 2004 , 1 davidson , k. 2005 , asp conf . ser . 332 : the fate of the most massive stars , 332 , 103 davidson , k. , dufour , r. j. , walborn , n. r. , & gull , t. r. 1986 , , 305 , 867 davidson , k. , ebbets , d. , weigelt , g. , humphreys , r. m. , hajian , a. r. , walborn , n. r. , & rosa , m. 1995 , , 109 , 1784 davidson , k. , & humphreys , r. m. 1986 , , 164 , l7 davidson , k. , ebbets , d. , weigelt , g. , humphreys , r. m. , hajian , a. r. , walborn , n. r. , & rosa , m. 1995 , , 109 , 1784 davidson , k. , & humphreys , r. m. 1997 , , 35 , 1 davidson , k. , humphreys , r. m. , ishibashi , k. , gull , t. r. , hamuy , m. , berdnikov , l. , & whitelock , p. 1999a , , 7146 , 1 davidson , k. , gull , t. r. , humphreys , r. m. , ishibashi , k. , whitelock , p. , berdnikov , l. , mcgregor , p. j. , metcalfe , t. s. , polomski , e. , & hamuy , m. 1999b , , 118 , 1777 davidson , k. , ishibashi , k. , gull , t. r. , & humphreys , r. m. 1999c , asp conf . ser . 179 : eta carinae at the millennium , 179 , 227 davidson , k. , martin , j. , humphreys , r. m. , ishibashi , k. , gull , t. r. , stahl , o. , weis , k. , hillier , d. j. , damineli , a. , corcoran , m. , & hamann , f. 2005 , , 129 , 900 davidson , k. , & ruiz , m .- t .
1975 , , 202 , 421 de vaucouleurs , g. , & eggen , o. j. 1952 , , 64 , 185 downes , r. , clampin , m. , shaw , r. , baum , s. , kinney , e. & mcgrath , m. 1997,stis instrument science report 97 - 03b , ( baltimore : stsci ) feast , m. , whitelock , p. , & marang , f. 2001 , , 322 , 741 feinstein , a. , & marraco , h. g. 1974 , , 30 , 271 feinstein , a. 1967 , the observatory , 87 , 287 fernandez lajus , e. , gamen , r. , schwartz , m. , salerno , n. , llinares , c. , farina , c. , amorn , r. , & niemela , v. 2003 , informational bulletin on variable stars , 5477 , 1 frew , d. j. 2005 , asp conf .
ser . 332 : the fate of the most massive stars , 332 , 158 gaviola , e. 1950 , , 111 , 408 gaviola , e. 1953 , , 118 , 234 gehrz , r.d . , & ney , e.p .
1972 , _ sky & tel . , _ 44 , 4 gonzaga , s. , et al .
2005,_acs instrument handbook , _ version 6.0 ( baltimore : stsci ) gull , t. r. , davidson , k. , & ishibashi , k. 2000 , american institute of physics conference series , 522 , 439 hofmann , k .- h . , & weigelt , g. 1988 , , 203 , l21 humphreys , r. m. , & davidson , k. 1994 , , 106 , 1025 humphreys , r. m. , & koppelman , m. 2005 , asp conf .
ser . 332 : the fate of the most massive stars , 332 , 159 kim quijano , j. , et al . 2003 , stis instrument handbook , version 7.0 ( baltimore : stsci ) koornneef , j. , bohlin , r. , buser , r. , horne , k. , & turnshek , d. 1986 , highlights of astronomy , 7 , 833 mattei , j. & foster , g. 1998 , iappp comm . , 72 , 53 martin , j. c. , & koppelman , m. d. 2004 , , 127 , 2352 martin , j. c. 2005 , asp conf . ser . 332 : the fate of the most massive stars , 332 , 111 martin , j. c. , davidson , k. , humphreys , r. m. , hillier , d. j. , & ishibashi , k. 2006a , , 640 , 474 martin , j. c. , davidson , k. , hamann , f. , stahl , o. , & weis , k. 2006 , , 118 , 697 oconnell , d. j. k. 1956 , _ vistas in astronomy , _ 2 , 1165 shaviv , n. j. 2005 , asp conf .
ser . 332 : the fate of the most massive stars , 332 , 180 sirianni , m. , et al . 2005 , , 117 , 1049 smith , n. , davidson , k. , gull , t. r. , ishibashi , k. , & hillier , d. j. 2003 , , 586 , 432 sterken , c. , freyhammer , l. , arentoft , t. , & van genderen , a. m. 1999 , , 346 , l33 sterken , c. , et al .
2001 , asp conf . ser . 233 : p cygni 2000 : 400 years of progress , 233 , 71 thackeray , a. d. 1953 , , 113 , 237 van genderen , a. m. , & th , p. s. 1984 , space sci .
, 39 , 317 van genderen , a. m. , sterken , c. , de groot , m. , & burki , g. 1999 , , 343 , 847 van genderen , a. m. , sterken , c. , & allen , w. h. 2003 , , 405 , 1057 viotti , r. , rossi , l. , cassatella , a. , altamore , a. , & baratta , g. b. 1989 , , 71 , 983 weigelt , g. , & ebersberger , j. 1986 , , 163 , l5 whitelock , p. a. , feast , m. w. , koen , c. , roberts , g. , & carter , b. s. 1994 , , 270 , 364 whitelock , p. a. , feast , m. w. , marang , f. , & breedt , e. 2004 , , 352 , 447 zanella , r. , wolf , b. , & stahl , o. 1984 , , 137 , 79 carinae .
the bottom panel shows the total relative response of each ccd and filter combination used in this study on the same wavelength scale as the top panel . for plotting purposes the curves
are not representative of relative responses between filters .
stis filters are plotted with a dashed line and acs / hrc filters are plotted with a solid line .
the dotted line represents the product of the stis ccd+f25nd3 response curve and the photon flux from the central star . ]
lcccc acs / hrc aperture correction & [email protected]&[email protected]&[email protected]&[email protected] + acs / hrc effective bandwidth & 187.29&239.41&173.75&165.20 + stis throughput correction & 1.6318&1.2903&0.6003&0.5601 + lllrrrr + + j8ma7ac7q&53345.391&2004.931&5.0&0.392&7.417&[email protected] + j8ma7acbq&53345.395&2004.931&5.0&0.391&7.418 & + j8ma7acfq&53345.398&2004.931&5.0&0.389&7.424 & + j8ma7acrq&53345.434&2004.931&5.0&0.405&7.382 & + j8ma8aorq&53565.266&2005.534&5.0&0.462&7.239&[email protected] + j8ma8ap4q&53565.301&2005.534&5.0&0.472&7.216 & + j8ma9aetq&53680.328&2005.849&5.0&0.467&7.226&[email protected] + j8ma9af1q&53680.344&2005.849&5.0&0.462&7.239 & + j8ma9afaq&53680.355&2005.849&5.0&0.465&7.231 & + j8ma9afnq&53680.371&2005.849&5.0&0.468&7.225 & + j9p602req&53951.121&2006.591&4.0&0.496&7.162&[email protected] + j9p602rhq&53951.125&2006.591&4.0&0.493&7.169 & + j9p602rkq&53951.125&2006.591&4.0&0.494&7.165 & + j9p602rnq&53951.129&2006.591&4.0&0.493&7.168 & + + + + j8ma7ac8q&53345.391&2004.931&1.4&0.939&6.469&[email protected] + j8ma7accq&53345.395&2004.931&1.4&0.941&6.466 & + j8ma7achq&53345.402&2004.931&1.4&0.928&6.481 & + j8ma7actq&53345.438&2004.931&1.4&0.967&6.437 & + j8ma8aouq&53565.273&2005.534&1.4&1.029&6.369&[email protected] + j8ma8ap7q&53565.309&2005.534&1.4&1.060&6.337 & + j8ma9aewq&53680.336&2005.849&1.4&1.126&6.271&[email protected] + j8ma9af5q&53680.348&2005.849&1.4&1.119&6.277 & + j8ma9afqq&53680.348&2005.849&1.4&1.137&6.261 & + j9p601qnq&53951.051&2006.591&1.0&1.261&6.148&[email protected] + j9p601qqq&53951.055&2006.591&1.0&1.264&6.145 & + j9p601qtq&53951.059&2006.591&1.0&1.258&6.151 & + j9p601qwq&53951.063&2006.591&1.0&1.256&6.152 & + + + + j8ma7ac9q&53345.391&2004.931&0.8&1.150&6.248&[email protected] + j8ma7acdq&53345.395&2004.931&0.8&1.140&6.258 & + j8ma7aclq&53345.406&2004.931&0.8&1.142&6.256 & + j8ma7acxq&53345.441&2004.931&0.8&1.190&6.211 & + j8ma8aoyq&53565.277&2005.534&0.8&1.190&6.211&[email protected] + j8ma8apbq&53565.313&2005.534&0.8&1.213&6.190 & + j8ma9aezq&53680.340&2005.849&0.8&1.440&6.004&[email protected] + j8ma9af8q&53680.352&2005.849&0.8&1.440&6.004 & + j8ma9afhq&53680.367&2005.849&0.8&1.441&6.003 & + j8ma9afuq&53680.383&2005.849&0.8&1.452&5.995 & + j9p601qxq&53951.066&2006.591&0.2&1.671&5.843&[email protected] + j9p601qzq&53951.066&2006.591&0.2&1.688&5.831 & + j9p601r0q&53951.066&2006.591&0.2&1.666&5.846 & + j9p602roq&53951.129&2006.591&0.3&1.638&5.864 & + j9p602rpq&53951.133&2006.591&0.3&1.656&5.852 & + j9p602rqq&53951.133&2006.591&0.3&1.643&5.861 & + j9p602rrq&53951.133&2006.591&0.3&1.653&5.854 & + + + + j8ma7acaq&53345.395&2004.931&0.1&1.337&6.084&[email protected] + j8ma7aceq&53345.398&2004.931&0.1&1.324&6.095 & + j8ma7acoq&53345.410&2004.931&0.1&1.318&6.100 & + j8ma7ad0q&53345.445&2004.931&0.1&1.370&6.059 & + j8ma8ap1q&53565.281&2005.534&0.1&1.342&6.081&[email protected] + j8ma8apeq&53565.281&2005.534&0.1&1.351&6.073 & + j8ma9af0q&53680.340&2005.849&0.1&1.589&5.897&[email protected] + j8ma9af9q&53680.355&2005.849&0.1&1.582&5.902 & + j8ma9afkq&53680.371&2005.849&0.1&1.606&5.886 & + j8ma9afxq&53680.387&2005.849&0.1&1.614&5.880 & + j9p602ryq&53951.145&2006.591&0.1&1.812&5.755&[email protected] + j9p602s0q&53951.145&2006.591&0.1&1.798&5.763 & + j9p602s2q&53951.145&2006.591&0.1&1.819&5.750 & + j9p602s4q&53951.148&2006.591&0.1&1.801&5.761 & + llrrrr o4j8010y0&50891.6 & -28&g230mb&1854&456.0 + o4j8010o0&50891.6 & -28&g230mb&1995&360.0 + o4j8011b0&50891.7 & -28&g230mb&2135&180.0 + o4j8011c0&50891.7 & -28&g230mb&2276&180.0 + o4j8011a0&50891.7 & -28&g230mb&2416&240.0 + o4j801040&50891.4 & -28&g230mb&2557&300.0 + o4j8010e0&50891.5 & -28&g230mb&2697&290.0 + o4j801050&50891.4 & -28&g230mb&2836&330.0 + o4j8010f0&50891.5 & -28&g230mb&2976&348.0 + o4j8010g0&50891.5 & -28&g230mb&3115&336.0 + o4j8010h0&50891.5 & -28&g430m&3165&144.0 + o4j801170&50891.7 & -28&g430m&3423 & 60.0 + o4j801180&50891.7 & -28&g430m&3680 & 72.0 + o4j8010z0&50891.6 & -28&g430m&3936 & 72.0 + o4j801060&50891.4 & -28&g430m&4194 & 36.0 + o4j8010x0&50891.6 & -28&g430m&4961 & 36.0 + o4j8010i0&50891.5 & -28&g430m&5216 & 36.0 + o4j801190&50891.7 & -28&g430m&5471 & 36.0 + o4j8010d0&50891.5 & -28&g750m&5734 & 15.0 + o4j801120&50891.7 & -28&g750m&6252 & 9.4 + o556020t0&51230.6 & -28&g230mb&1854&380.0 + o556020k0&51230.5 & -28&g230mb&1995&278.0 + o55602110&51230.7 & -28&g230mb&2135&150.0 + o55602120&51230.7 & -28&g230mb&2276&150.0 + o55602100&51230.7 & -28&g230mb&2416&200.0 + o556020b0&51230.5 & -28&g230mb&2557&300.0 + o556020e0&51230.5 & -28&g230mb&2697&280.0 + o55602040&51230.5 & -28&g230mb&2836&307.0 + o556020z0&51230.6 & -28&g230mb&2976&360.0 + o556020c0&51230.5 & -28&g230mb&3115&280.0 + o556020v0&51230.6 & -28&g430m&3165&120.0 + o556020f0&51230.5 & -28&g430m&3423 & 50.0 + o556020d0&51230.5 & -28&g430m&3680 & 60.0 + o556020s0&51230.6 & -28&g430m&3936 & 60.0 + o556020y0&51230.6 & -28&g430m&4961 & 30.0 + o556020g0&51230.5 & -28&g430m&5216 & 30.0 + o556020w0&51230.6 & -28&g430m&5471 & 30.0 + o55602090&51230.5 & -28&g750m&5734 & 15.0 + o556020p0&51230.6 & -28&g750m&6252 & 8.0 + o62r010p0&52016.9 & 22&g230mb&1854&340.0 + o62r010i0&52016.8 & 22&g230mb&1995&300.0 + o62r010m0&52016.8 & 22&g230mb&2135&280.0 + o62r010x0&52016.9 & 22&g230mb&2276&300.0 + o62r010y0&52016.9 & 22&g230mb&2416&200.0 + o62r010a0&52016.8 & 22&g230mb&2557&420.0 + o62r010c0&52016.8 & 22&g230mb&2697&280.0 + o62r01040&52016.8 & 22&g230mb&2836&300.0 + o62r010t0&52016.9 & 22&g230mb&2976&360.0 + o62r01080&52016.8 & 22&g230mb&3115&350.0 + o62r010r0&52016.9 & 22&g430m&3165&100.0 + o62r010d0&52016.8 & 22&g430m&3423 & 70.0 + o62r010b0&52016.8 & 22&g430m&3680 & 60.0 + o62r010o0&52016.8 & 22&g430m&3936 & 32.0 + o62r01030&52016.7 & 22&g430m&4194 & 30.0 + o62r010v0&52016.9 & 22&g430m&4961 & 36.0 + o62r010e0&52016.8 & 22&g430m&5216 & 14.0 + o62r010s0&52016.9 & 22&g430m&5471 & 30.0 + o62r01090&52016.8 & 22&g750m&5734 & 8.0 + o62r010l0&52016.8 & 22&g750m&6252 & 10.0 + o6ex030e0&52183.2 & 165&g430m&4961 & 36.0 + o6ex030c0&52183.1 & 165&g430m&5216 & 36.0 + o6ex030b0&52183.1 & 165&g750m&5734 & 15.0 + o6ex02080&52294.0 & -82&g230mb&1854&800.0 + o6ex020i0&52294.1 & -82&g230mb&1995&600.0 + o6ex020m0&52294.1 & -82&g230mb&2135&600.0 + o6ex020x0&52294.2 & -82&g230mb&2276&600.0 + o6ex020y0&52294.2 & -82&g230mb&2416&320.0 + o6ex020a0&52294.0 & -82&g230mb&2557&1200.0 + o6ex020c0&52294.1 & -82&g230mb&2697&280.0 + o6ex02040&52294.0 & -82&g230mb&2836&300.0 + o6ex020t0&52294.1 & -82&g230mb&2976&340.0 + o6ex020p0&52294.1 & -82&g230mb&3115&300.0 + o6ex020r0&52294.1 & -82&g430m&3165 & 90.0 + o6ex020d0&52294.1 & -82&g430m&3423 & 90.0 + o6ex020b0&52294.1 & -82&g430m&3680 & 52.0 + o6ex020o0&52294.1 & -82&g430m&3936 & 26.0 + o6ex02030&52294.0 & -82&g430m&4194 & 18.0 + o6ex020v0&52294.2 & -82&g430m&4961 & 36.0 + o6ex020e0&52294.1 & -82&g430m&5216 & 16.0 + o6ex020s0&52294.1 & -82&g430m&5471 & 34.0 + o6ex02090&52294.0 & -82&g750m&5734 & 6.0 + o6ex020l0&52294.1 & -82&g750m&6252 & 8.0 + o6mo020a0&52459.5 & 69&g230mb&1854&400.0 + o6mo020x0&52459.6 & 69&g230mb&1995&300.0 + o6mo02120&52459.6 & 69&g230mb&2135&300.0 + o6mo021n0&52459.7 & 69&g230mb&2276&300.0 + o6mo021e0&52459.7 & 69&g230mb&2416&320.0 + o6mo020h0&52459.6 & 69&g230mb&2557&400.0 + o6mo020m0&52459.6 & 69&g230mb&2697&340.0 + o6mo02050&52459.5 & 69&g230mb&2836&300.0 + o6mo021i0&52459.7 & 69&g230mb&2976&320.0 + o6mo02190&52459.7 & 69&g230mb&3115&300.0 + o6mo021r0&52459.7 & 69&g430m&3165 & 90.0 + o6mo020p0&52459.6 & 69&g430m&3423 & 90.0 + o6mo020l0&52459.6 & 69&g430m&3680 & 52.0 + o6mo021a0&52459.7 & 69&g430m&3936 & 26.0 + o6mo02060&52459.5 & 69&g430m&4194 & 18.0 + o6mo021m0&52459.7 & 69&g430m&4961 & 36.0 + o6mo020q0&52459.6 & 69&g430m&5216 & 16.0 + o6mo021h0&52459.7 & 69&g430m&5471 & 34.0 + o6mo020i0&52459.6 & 69&g750m&5734 & 9.0 + o6mo02150&52459.7 & 69&g750m&6252 & 8.0 + o8gm12060&52682.9 & -57&g230mb&1854&600.0 + o8gm120h0&52683.0 & -57&g230mb&1995&600.0 + o8gm120l0&52683.0 & -57&g230mb&2135&600.0 + o8gm120w0&52683.0 & -57&g230mb&2276&600.0 + o8gm120r0&52683.0 & -57&g230mb&2416&320.0 + o8gm12090&52682.9 & -57&g230mb&2557&800.0 + o8gm120c0&52682.9 & -57&g230mb&2697&340.0 + o8gm12030&52682.9 & -57&g230mb&2836&300.0 + o8gm120t0&52683.0 & -57&g230mb&2976&340.0 + o8gm120o0&52683.0 & -57&g230mb&3115&300.0 + o8gm120n0&52683.0 & -57&g430m&3165 & 90.0 + o8gm120d0&52682.9 & -57&g430m&3423 & 90.0 + o8gm120b0&52682.9 & -57&g430m&3680 & 52.0 + o8gm120p0&52683.0 & -57&g430m&3936 & 26.0 + o8gm12040&52682.9 & -57&g430m&4194 & 18.0 + o8gm120v0&52683.0 & -57&g430m&4961 & 36.0 + o8gm120e0&52682.9 & -57&g430m&5216 & 16.0 + o8gm120s0&52683.0 & -57&g430m&5471 & 34.0 + o8gm120a0&52682.9 & -57&g750m&5734 & 6.0 + o8gm12050&52682.9 & -57&g750m&6252 & 8.0 + o8gm210g0&52727.3 & -28&g430m&4961 & 16.0 + o8gm210e0&52727.3 & -28&g430m&5216 & 16.0 + o8gm210b0&52727.3 & -28&g430m&5471 & 34.0 + o8gm210d0&52727.3 & -28&g750m&5734 & 9.0 + o8gm410g0&52764.4 & 27&g430m&4961 & 16.0 + o8gm410e0&52764.4 & 27&g430m&5216 & 16.0 + o8gm410b0&52764.3 & 27&g430m&5471 & 34.0 + o8gm410d0&52764.3 & 27&g750m&5734 & 9.0 + o8gm320a0&52778.5 & 38&g230mb&1854&400.0 + o8gm320x0&52778.9 & 38&g230mb&1995&300.0 + o8gm33020&52776.4 & 38&g230mb&2135&300.0 + o8gm330n0&52776.6 & 38&g230mb&2276&300.0 + o8gm330e0&52776.5 & 38&g230mb&2416&320.0 + o8gm320h0&52778.6 & 38&g230mb&2557&400.0 + o8gm320m0&52778.7 & 38&g230mb&2697&340.0 + o8gm32050&52778.5 & 38&g230mb&2836&300.0 + o8gm330i0&52776.5 & 38&g230mb&2976&320.0 + o8gm33090&52776.5 & 38&g230mb&3115&300.0 + o8gm33060&52776.4 & 38&g430m&3165 & 90.0 + o8gm320p0&52778.8 & 38&g430m&3423 & 90.0 + o8gm320l0&52778.7 & 38&g430m&3680 & 52.0 + o8gm330a0&52776.5 & 38&g430m&3936 & 26.0 + o8gm32060&52778.5 & 38&g430m&4194 & 18.0 + o8gm330m0&52776.6 & 38&g430m&4961 & 36.0 + o8gm320q0&52778.8 & 38&g430m&5216 & 16.0 + o8gm330h0&52776.5 & 38&g430m&5471 & 34.0 + o8gm320i0&52778.6 & 38&g750m&5734 & 9.0 + o8gm330r0&52776.6 & 38&g750m&6252 & 8.0 + o8gm520a0&52791.7 & 62&g230mb&1854&400.0 + o8gm520x0&52791.8 & 62&g230mb&1995&300.0 + o8gm52100&52791.9 & 62&g230mb&2135&400.0 + o8gm521o0&52792.0 & 62&g230mb&2276&300.0 + o8gm521f0&52791.9 & 62&g230mb&2416&320.0 + o8gm520h0&52791.8 & 62&g230mb&2557&400.0 + o8gm520m0&52791.8 & 62&g230mb&2697&340.0 + o8gm52050&52791.7 & 62&g230mb&2836&300.0 + o8gm521j0&52791.9 & 62&g230mb&2976&340.0 + o8gm52180&52791.9 & 62&g230mb&3115&300.0 + o8gm52170&52791.9 & 62&g430m&3165 & 90.0 + o8gm520p0&52791.8 & 62&g430m&3423 & 90.0 + o8gm520l0&52791.8 & 62&g430m&3680 & 52.0 + o8gm521b0&52791.9 & 62&g430m&3936 & 26.0 + o8gm52060&52791.7 & 62&g430m&4194 & 18.0 + o8gm521k0&52791.9 & 62&g430m&4961 & 36.0 + o8gm520q0&52791.8 & 62&g430m&5216 & 16.0 + o8gm521g0&52791.9 & 62&g430m&5471 & 34.0 + o8gm520i0&52791.8 & 62&g750m&5734 & 9.0 + o8gm521s0&52792.0 & 62&g750m&6252 & 8.0 + o8gm620a0&52813.8 & 70&g230mb&1854&400.0 + o8gm620x0&52814.2 & 70&g230mb&1995&300.0 + o8gm62100&52814.2 & 70&g230mb&2135&300.0 + o8gm630d0&52812.2 & 70&g230mb&2276&300.0 + o8gm63040&52812.1 & 70&g230mb&2416&350.0 + o8gm620h0&52814.0 & 70&g230mb&2557&400.0 + o8gm620m0&52814.1 & 70&g230mb&2697&340.0 + o8gm62050&52813.8 & 70&g230mb&2836&300.0 + o8gm63080&52812.2 & 70&g230mb&2976&320.0 + o8gm62170&52814.3 & 70&g230mb&3115&260.0 + o8gm62140&52814.3 & 70&g430m&3165 & 90.0 + o8gm620p0&52814.1 & 70&g430m&3423 & 90.0 + o8gm620l0&52814.1 & 70&g430m&3680 & 52.0 + o8gm62180&52814.3 & 70&g430m&3936 & 26.0 + o8gm62060&52813.8 & 70&g430m&4194 & 18.0 + o8gm630c0&52812.2 & 70&g430m&4961 & 36.0 + o8gm620q0&52814.1 & 70&g430m&5216 & 16.0 + o8gm63070&52812.1 & 70&g430m&5471 & 34.0 + o8gm620i0&52814.0 & 70&g750m&5734 & 9.0 + o8gm630h0&52812.2 & 70&g750m&6252 & 15.0 + o8ma720q0&52825.5 & 69&g430m&4961 & 16.0 + o8ma720p0&52825.5 & 69&g430m&5216 & 16.0 + o8ma720h0&52825.4 & 69&g750m&5734 & 9.0 + o8ma820b0&52852.0 & 105&g230mb&1854&400.0 + o8ma820y0&52852.2 & 105&g230mb&1995&300.0 + o8ma82110&52852.2 & 105&g230mb&2135&300.0 + o8ma821m0&52852.4 & 105&g230mb&2276&300.0 + o8ma821b0&52852.3 & 105&g230mb&2416&320.0 + o8ma820i0&52852.0 & 105&g230mb&2557&400.0 + o8ma820n0&52852.1 & 105&g230mb&2697&340.0 + o8ma82060&52851.9 & 105&g230mb&2836&300.0 + o8ma821i0&52852.4 & 105&g230mb&2976&300.0 + o8ma82160&52852.3 & 105&g230mb&3115&300.0 + o8ma821a0&52852.3 & 105&g430m&3165 & 90.0 + o8ma820q0&52852.1 & 105&g430m&3423 & 90.0 + o8ma820m0&52852.1 & 105&g430m&3680 & 52.0 + o8ma821e0&52852.3 & 105&g430m&3936 & 26.0 + o8ma82070&52851.9 & 105&g430m&4194 & 18.0 + o8ma821o0&52852.4 & 105&g430m&4961 & 32.0 + o8ma820r0&52852.1 & 105&g430m&5216 & 16.0 + o8ma821j0&52852.4 & 105&g430m&5471 & 34.0 + o8ma820j0&52852.1 & 105&g750m&5734 & 6.0 + o8ma820a0&52852.0 & 105&g750m&6252 & 8.0 + o8ma92070&52904.3 & 153&g230mb&1854&600.0 + o8ma920i0&52904.4 & 153&g230mb&1995&600.0 + o8ma920m0&52904.4 & 153&g230mb&2135&600.0 + o8ma920x0&52904.5 & 153&g230mb&2276&300.0 + o8ma920s0&52904.4 & 153&g230mb&2416&600.0 + o8ma920a0&52904.3 & 153&g230mb&2557&800.0 + o8ma920d0&52904.4 & 153&g230mb&2697&340.0 + o8ma92040&52904.3 & 153&g230mb&2836&300.0 + o8ma920u0&52904.5 & 153&g230mb&2976&340.0 + o8ma920p0&52904.4 & 153&g230mb&3115&300.0 + o8ma920o0&52904.4 & 153&g430m&3305 & 90.0 + o8ma920e0&52904.4 & 153&g430m&3423 & 90.0 + o8ma920c0&52904.4 & 153&g430m&3680 & 52.0 + o8ma920q0&52904.4 & 153&g430m&3936 & 26.0 + o8ma92050&52904.3 & 153&g430m&4194 & 18.0 + o8ma920w0&52904.5 & 153&g430m&4961 & 36.0 + o8ma920f0&52904.4 & 153&g430m&5216 & 16.0 + o8ma920t0&52904.5 & 153&g430m&5471 & 34.0 + o8ma920b0&52904.4 & 153&g750m&5734 & 6.0 + o8ma920z0&52904.5 & 153&g750m&6252 & 8.0 + o8ma830g0&52960.7&-142&g430m&4961 & 18.0 + o8ma830e0&52960.6&-142&g430m&5216 & 16.0 + o8ma830b0&52960.6&-142&g430m&5471 & 32.0 + o8ma830d0&52960.6&-142&g750m&5734 & 8.0 + o8ma940g0&53071.3 & -28&g230mb&1854&430.0 + o8ma940i0&53071.3 & -28&g230mb&2135&320.0 + o8ma940m0&53071.3 & -28&g230mb&2416&450.0 + o8ma94070&53071.3 & -28&g230mb&2557&410.0 + o8ma940e0&53071.3 & -28&g230mb&2697&323.0 + o8ma94020&53071.2 & -28&g230mb&2836&320.0 + o8ma940s0&53071.3 & -28&g230mb&2976&323.0 + o8ma940j0&53071.3 & -28&g230mb&3115&255.0 + o8ma940h0&53071.3 & -28&g430m&3165 & 90.0 + o8ma940a0&53071.3 & -28&g430m&3423 & 90.0 + o8ma94090&53071.3 & -28&g430m&3680 & 52.0 + o8ma940k0&53071.3 & -28&g430m&3936 & 26.0 + o8ma94030&53071.2 & -28&g430m&4194 & 18.0 + o8ma940p0&53071.3 & -28&g430m&4961 & 36.0 + o8ma940b0&53071.3 & -28&g430m&5216 & 16.0 + o8ma940n0&53071.3 & -28&g430m&5471 & 34.0 + o8ma94080&53071.3 & -28&g750m&5734 & 6.0 + o8ma940r0&53071.3 & -28&g750m&6252 & 10.0 + rrrr + + 50891.7&1998.21&0.076&9.194 + 51230.6&1999.14&0.155&8.425 + 52016.9&2001.29&0.189&8.211 + 52294.2&2002.05&0.230&7.996 + 53459.7&2002.51&0.255&7.885 + 52683.0&2003.12&0.263&7.850 + 52777.6&2003.37&0.224&8.025 + 52791.9&2003.41&0.239&7.953 + 52813.1&2003.47&0.202&8.136 + 52852.2&2003.58&0.126&8.652 + 52904.4&2003.72&0.132&8.599 + 53071.3&2004.18&0.218&8.053 + + + + 50891.6&1998.21&0.167&8.346 + 51230.6&1999.14&0.403&7.387 + 52016.9&2001.29&0.448&7.273 + 52294.2&2002.05&0.550&7.048 + 53459.7&2002.51&0.592&6.969 + 52683.0&2003.12&0.535&7.080 + 52777.6&2003.37&0.498&7.156 + 52791.9&2003.41&0.546&7.058 + 52813.1&2003.47&0.478&7.202 + 52852.2&2003.58&0.326&7.618 + 52904.4&2003.72&0.346&7.552 + 53071.3&2004.18&0.687&6.807 + + + + 50891.6&1998.21&0.237&7.964 + 51230.6&1999.14&0.516&7.119 + 52016.9&2001.29&0.524&7.101 + 52294.1&2002.05&0.625&6.910 + 53459.7&2002.51&0.626&6.908 + 52683.0&2003.12&0.531&7.088 + 52777.6&2003.37&0.553&7.043 + 52791.9&2003.41&0.619&6.920 + 52813.1&2003.47&0.644&6.878 + 52852.2&2003.58&0.572&7.007 + 52904.4&2003.72&0.653&6.862 + 53071.3&2004.18&0.894&6.521 + + + + 50891.6&1998.21&0.352&7.533 + 51230.6&1999.14&0.605&6.945 + 52016.9&2001.29&0.691&6.802 + 52183.2&2001.75&0.823&6.611 + 52294.1&2002.05&0.703&6.782 + 53459.7&2002.51&0.711&6.771 + 52683.0&2003.12&0.668&6.838 + 52727.3&2003.24&0.697&6.792 + 52764.4&2003.34&0.688&6.807 + 52777.6&2003.37&0.732&6.738 + 52791.9&2003.41&0.754&6.707 +
52813.1&2003.47&0.849&6.578 + 52825.5&2003.51&0.906&6.507 + 52852.2&2003.58&0.851&6.575 + 52904.5&2003.72&1.015&6.384 + 52960.6&2003.88&1.124&6.273 + 53071.3&2004.18&1.185&6.215 + | during the past decade @xmath0 car has brightened markedly , possibly indicating a change of state . here
we summarize photometry gathered by the hubble space telescope as part of the hst treasury project on this object .
our data include stis / ccd acquisition images , acs / hrc images in four filters , and synthetic photometry in flux - calibrated stis spectra .
the hst s spatial resolution allows us to examine the central star separate from the bright circumstellar ejecta . _
its apparent brightness continued to increase briskly during 200206 , especially after the mid-2003 spectroscopic event . _
if this trend continues , the central star will soon become brighter than its ejecta , quite different from the state that existed only a few years ago .
one precedent may be the rapid change observed in 19381953 .
we conjecture that the star s mass - loss rate has been decreasing throughout the past century . |
among the many theories which try to explain the features of the economic development of countries ( see for example @xcite ) , a recent , data driven and innovative approach is the one known as _ economic complexity_.
this approach is based on the country - product matrix @xmath2 , whose binary elements indicates if the given country exports a certain products or not ( in particular , if the country shows a revealed comparative advantage @xcite in that market ) . in contrast with the ricardian specialization paradigm @xcite , the @xmath2 matrix is , when suitably ordered , not block diagonal but triangular : this means that ubiquitous products are produced by all countries , while rare products are produced only by diversified countries . in a first attempt to use this information to evaluate the goodness of countries and products , hidalgo and
hausmann @xcite proposed an algorithm on the matrix @xmath2 .
tacchella and al .
@xcite deeply revisited the algorithm and proposed a new metrics to estimate the country s potential of growth , called the _ fitness f _ , in terms of the quality , or _
complexity q _ , of its exported products .
see @xcite for a comparison between the two approaches .
@xmath3 @xmath4 @xmath5 @xmath6 where @xmath7 is the fitness of country @xmath8 and @xmath9 is the complexity of product @xmath10 .
all the values are determined only by the country - product matrix , @xmath2 .
we point out that , obviously , in principle it is possible to apply this algorithm to any bipartite network defined by the matrix @xmath2 , so the definitions of countries and products as the rows and columns of the matrix is someway reductive .
however , since its first and main use so far is this one , in the following we will use the expressions country and product to denote the row and column elements . in this short note
we analyze the characteristics of @xmath2 required to have convergence of the algorithm to numbers strictly greater than zero , for both every @xmath7 and @xmath9 .
we will first show in section [ sec : theoretical ] a class of @xmath2 for which it is possible to solve analytically the algorithm in closed form and find these characteristics ; we therefore produce an _ ansatz _ for the general case .
we will then show in section [ sec : numerical ] that the insight is numerically supported for any analyzed random matrix and in section [ sec : real ] we will show how this is valid also for the matrices used in economic complexity . in the conclusions , in section [ sec : conclusions ] , we will explain the effect of this result on the economic complexity field and we will give some practical indications for the use of the algorithm in light of our findings .
finally there will be an additional appendix , [ sec : appendixgamma ] , in which we will try to generalize our findings to a wider class of algorithms similar to the one we have studied .
we will consider a specific class of matrices .
an example of the class is presented in equation [ eq : matrix ] . the generic matrix in the class is composed of 4 blocks , one of which of zeros .
the other three blocks will not be composed only of zeros but they will have some @xmath11s . the density of @xmath11s in the block can be in principle any value between @xmath12 and @xmath11 , but to allow for a closed form solution each block has to have the same number of @xmath11s and @xmath12s in all its columns and rows .
it is interesting to note that , since the algorithm holds for reordering of rows and columns , a matrix is also a member of the class if it is possible to reorder its rows and columns in such a way to obtain a matrix in the class .
we will number the four blocks from 1 to 4 anti - clockwise starting from the top right corner . in the following
we will always assume that the block 4 , the bottom right block , is the block with only zeros .
alternatively , will also refers to the block 4 as the _
external area_. we will also call the blocks 1 and 3 the _ frontier _ , block 2 the _ internal area_. in the following example of this class of matrices the four densities are , anti - clockwise from the top right corner , @xmath13 , @xmath11 , @xmath14 , @xmath12 .
@xmath15 the key variables will be the sizes of the four blocks : their height @xmath16 and @xmath17 and their width @xmath18 and @xmath19
. it will be revealed instead as almost unimportant the density of @xmath11s in the non - empty blocks . to make the notation easier in the following
we will assume that all the densities of the three non - empty blocks are equal to @xmath11 : the three non empty blocks ( 1 , 2 and 4 ) will be composed only of @xmath11s . in subsection
[ sec : density ] we will explain what changes when we relax this assumption .
the symmetry of the matrix is particularly useful because we can state , for symmetry reasons , that all the countries in the @xmath17 upper rows have the same fitness , and the same is true for the countries in the @xmath16 lower rows .
this is obvious since their rows are equal or can be made equal with a simple rearrangement of the columns .
the same is also true for the complexity of the first @xmath18 products and the complexity of the last @xmath19 products .
therefore the fitness at the iteration @xmath20 of the algorithm can be represented with just two numbers , with an obvious change of notation @xmath21 it is easy to compute the next iteration of complexities from equation [ eq : complexity ] , @xmath22 since we are not interested in the complexities , to avoid unnecessary computations we will not normalize accordingly to [ eq : complexitynorm ] . in the next steps we will compute the fitness values and we will normalize them : any multiplicative factor to all the complexities
would be removed in any case at that step .
thanks to this property we can rewrite the complexities in an easier way for the next computation , in particular as @xmath23 where @xmath24 is proportional to @xmath25 . we can now compute the iteration @xmath26 for the fitness from equation [ eq : fitness ] . since we are using @xmath24 instead of @xmath1 ,
the values obtained by equation [ eq : fitness ] will be proportional to @xmath27 . defining these as @xmath28
, we obtain @xmath29 where the second step is due to imposing the normalization condition : @xmath30 this constraint allows us to reduce the number of variables from 2 to 1 and to derive a close solution . finally , normalizing accordingly to equation [ eq : fitnessnorm ] we have @xmath31 let s focus only at the second component of the fitness vector , which in the following we will call @xmath32 , and which is referred to the fitnesses of the @xmath16 lowest rows . comparing [ eq : definition ] and [ eq : nextstep ]
we see that , in one algorithmic step , we had @xmath33 where @xmath34 from which the closed form solution , by induction , is trivial @xmath35 the solution for the other component can be obtained from the normalization condition given by eq .
[ eq : normalization ] how does @xmath32 behave when @xmath20 goes to infinity ?
it trivially depends on @xmath36 : * if @xmath37 , @xmath38 converges to @xmath12 exponentially fast while @xmath20 goes to infinity ; * if @xmath39 , @xmath40 and therefore converges to @xmath12 as fast as @xmath41 ; * if @xmath42 , @xmath38 converges exponentially to a non zero number , @xmath43 . from [ eq : a2 ] we notice now that @xmath36 has also a particular geometric meaning : it is the ratio of the areas of the bottom right block ( block 4 ) and the top left block ( block 2 ) . remembering our definitions of the blocks , @xmath36 is therefore greater than @xmath11 , and the second fitness component @xmath32 converges to @xmath12 , if the _ external area _ is bigger than the _ internal area_. or , visually , if the belly of the non - zeros area is inward .
@xmath39 , when @xmath36 is defined through equation [ eq : a2 ] , defines the diagonal line , or its obvious geometric extension to rectangular matrices , @xmath44 it is worth noting that if @xmath36 is close , but different , to @xmath11 , the exponential behavior of the fitness can be very slow .
in particular it is possible to see that , for @xmath37 , @xmath45 for @xmath36 near enough to @xmath11 and @xmath20 large enough .
this define a characteristic time @xmath46 of the exponential that can be very large when @xmath36 approaches @xmath11 .
similarly for @xmath42 , @xmath47 , again in the case of @xmath36 near enough to @xmath11 and @xmath20 large enough .
interestingly the result does not change if the densities of @xmath11s in the blocks 1 , 2 and 3 are different than @xmath11 , if the block 4 is still composed only of @xmath12s and the block 2 has a positive density only the shape matters , and not the density .
in particular doing the same computations with densities different from @xmath11 ( and potentially heterogeneous ) we find that : * equation [ eq : closed ] does not change , * @xmath36 does not change , * @xmath48 does depend on the densities of the blocks 1 and 3 , but not from the density of block 2 if it is not @xmath12 ( see ahead for the case in which the density of block 3 is zero ) .
the specific value of the density of @xmath11s and @xmath12s in the internal area is therefore completely irrelevant , while the density inside the frontier only determines the specific value of the convergence point ( if there is convergence to a number greater than @xmath12 ) .
the case in which the density in the internal area is @xmath12 is a peculiar case , since in this case it is not possible to really define an inside and an outside .
in fact it is possible to rearrange the rows and columns to switch inside and outside . in this case
it is still possible , causing @xmath36 to go to @xmath49 .
therefore @xmath36 is defined but for a possible multiplicative inversion . ] also to define @xmath36 . if @xmath36 is different than @xmath11 the fitness of one of the two sets of countries is always converging to zero . in the case
in which @xmath36 is equal @xmath11 instead , any value of fitness is a stationary point of the algorithm and therefore the fitnesses produced by the algorithm will be equal to the starting conditions . this can be said in a different way . in this case
the blocks 1 and 3 are not connected through a product common among countries belonging different blocks .
therefore the algorithm in this case does not have any direct information to compare fitnesses and complexities of countries and products in different blocks . to compare the two blocks it is therefore only possible to use their densities and shape .
however the density turns out to be irrelevant , since the more products in a block the more is diluted their influence on the fitnesses of the countries in their block : as it is possible to compute , this cancel out the additional effect of having more or less products . only the relative shape among the blocks , measured by @xmath36 ,
is left to compare the two blocks .
if @xmath36 is equal 1 , and therefore also the shape does not help to compare the blocks , any fitness and complexity is consistent . if @xmath36 is different than 1 , it will determine the convergence to @xmath12 of the fitnesses and complexities of one of the two blocks .
the ansatz that will be then corroborated by the numerical investigation in the next sections is therefore simple , at least using loose definitions : if the belly of the matrix is outward all the fitnesses and complexities converge to numbers greater than @xmath12 .
if the belly is inward , some of the fitnesses will converge to @xmath12 .
while in the case defined in [ sec : matrix ] the conversational definition of `` inward belly '' - corresponding to @xmath37 - is unambiguous , for the general cases that we will investigate in the sections [ sec : numerical ] and [ sec : real ] there are many possible definitions . defining * _ ordered _ matrix , as the matrix @xmath2 after rearranging the columns and rows such that all the countries are ranked accordingly to their fitnesses and all the products are ranked accordingly to their complexities for many countries , for any finite number of iteration the fitness values will be greater than @xmath12 . moreover ,
since the convergence speed after some iterations is constant for each country and products , there will be a number of iterations such that , for any following iteration , the ranking of fitnesses and complexities will be constant . ] ; * _ diagonal _ line , as the line , in the previously defined matrix , going from the least fit country and least complex product to the most fit country and most complex product ; if the matrix is squared , this definition is the usual definition of diagonal of the matrix ; if the matrix is rectangular , this definition is the trivial geometric extension of the previous one . *
the _ external _ area of the matrix is the joint set of zeros in the ordered matrix including the corner corresponding to the @xmath12 for the lowest fitness and highest complexity product ; we will see that a proper definition , with relative guess , is [ ansatz : belly ] a matrix @xmath2 have all the fitness and complexity different from zero if , after ordering said matrix , the diagonal line does not pass through the external area .
it is worth to note that when a country s fitness converges to zero , from equation [ eq : complexity ] , we know that all the exported products complexity will converge to zero too . at the same time , the effect on the fitnesses of countries which export a product whose complexity converges to zero , from equation [ eq : fitness ] , is negligible .
therefore , we can consider that , but for multiplicative common factors coming from equations [ eq : fitnessnorm ] and [ eq : complexitynorm ] , removing from the matrix @xmath2 a line corresponding to a country converging to zero fitness and all the columns corresponding to its exported products , the output will be not affected .
it is therefore reasonable to make a further guess : [ ansatz : cutpoint ] if the diagonal line does pass through the external part of a ordered matrix @xmath2 , some countries and products will converge to zero . if we progressively remove them from the analysis , defining new ordered matrices @xmath2 , we will have a convergence to finite values of fitnesses and complexities for the remaining countries and products for which the the new @xmath2 have a diagonal line not passing through the external area .
these ansatz implicitly define the _ crossing country _ of a matrix as the lowest fitness country that converges to a nonzero fitness .
in other words , it is the first country that does not need to be removed in the removal process described above , defining the largest matrix such that all its countries have nonzero fitness .
note that in general searching instead from the top countries does not give the proper result in a situation in which the diagonal cross multiple times the external area . if one starts the check from the top countries and keeps adding countries with lower fitnesses , the crossing country is not the first one met .
in this case of multiple crossing , the crossing country is the first country met in the process starting from the bottom , as in ansatz [ ansatz : cutpoint ] .
the guesses [ ansatz : belly ] and [ ansatz : cutpoint ] will be tested numerically in the sections [ sec : numerical ] and [ sec : real ] .
in this section we present the results of our numerical simulations , in which we study some peculiar behaviors of sample matrices which we did not treat analytically .
in particular , by means of synthetic examples of possible @xmath2 matrices we study the cases in which more than two blocks are present , the importance of producing rare products and we investigate the ansatz proposed in the previous section .
the reader will notice that the matrices investigated in the first two subsections correspond to the case @xmath39 discussed above . in the case of more of two blocks
it is not possible to write the components of @xmath50 in the form given by eq.[eq : definition ] , because one would have more than one unknown variable .
however , what one can do is progressively reduce the given matrix to submatrices , each one made of two blocks . in the following example we present a block diagonal matrix : @xmath51 we have found that if off block diagonal elements are absent ( that is , both the internal and the external areas are empty ) all countries will remain in the respective initial conditions . as we expected from the two block analysis presented in sec.[sec : density_zeros ] , this result is stable even if density is lower : only the relative dimensions of the blocks counts .
the presence of common products , in the very spirit of @xcite , does not add much information other than the fact that those countries which produce only very common products can not be very high in the fitness ranking . as a consequence
, we can expect _ monopolies _ of already diversified countries to be relevant . to investigate the consequences of having common ubiquitous products we take into account a synthetic situation in which three countries produce only very common products , two countries have a _
duopoly _ and one country , which has the same diversification of these two countries , has a monopoly .
the results are the following : @xmath52 on the left side of the matrix we show the speed of the fitness decays with the iteration @xmath20 , with _
c _ meaning convergence to a finite value of fitness . for example
, the notation @xmath53 means that the fitness @xmath54 of the country @xmath8 corresponding to that row is converging towards zero with a power law of the number @xmath20 of the algorithm iterations , that is , @xmath55 for a sufficiently larger @xmath20 .
as one can notice , the presence of a monopoly makes the first country the only one to converge to a nonzero value of @xmath0 , while the advantage given by the extra products makes the second and third country perform better with respect to the last three countries . in any case , this advantage is not striking , because we are still in the case in which @xmath39 , so the difference is only in the exponent of the decay .
we stress that the presence of a different decay means that there exist a number @xmath56 so that from iteration @xmath57 the ranking will stay constant , and so the ranking will be given by the decay exponent . +
a further , self - illustrative situation of the consequences of having common products is presented below .
@xmath58 let us now consider the following matrix : @xmath59 here we have three blocks of nations , with one monopoly ( country 3 ) and two duopolies ( countries 1 and 2 ) . on the basis of the previous considerations , we know that the 3x3 submatrix constituted by countries 1,2 and 3 and products 4 , 5 and 6 is degenerate , in the sense that every initial condition @xmath60 is stable under the iterations ( also the whole matrix would be degenerate if country 1 would not have produced product 2 ) .
however , in this case , the presence of an external product changes the situation , giving an advantage to the duopoly , which converges to a finite value of the fitness and makes the monopoly tend to zero , even if with a lower exponent ( @xmath61 instead of @xmath11 ) . by means of extensive numerical simulation
we have found that the specific value of the decay is given by the relative size of the worst block ( the one which decays with @xmath41 ) with respect with the total size of the externally connected block , that is , the sum of the sizes of the blocks which are connected by the external @xmath11 . in other words ,
the decay is given by the ratio between the worst block and the sum of the sizes of all the blocks but the one whose decay we are computing .
in the situation presented above , the worst block has size 3 and the converging block has size 2 , from which we have @xmath62 . to make another example we can imagine a different case , in which we give an advantage to the monopoly with respect to the block of size 3 ; in this case , the duopoly will decay with an exponent equal to @xmath63 . +
another interesting phenomenon is evident .
let us consider the set of countries 4 , 5 and 6 . even if the second product is owned only by countries 4 and 5 , also country 6 decays in an equally faster way .
this is due to the fact that the last block is _ connected _ , that is , there exist a path of products that connects all the countries in the block , and the absence of monopolies in the block .
until now we have seen that different power law decays come out from , in general , n - polistic competitions whose symmetry is broken by the presence of products external with respect to the competitors . in the case
in which one or more n - poly has a further n - poly which is not shared , the decay of the competitor will be _ exponential_.
here it is an example : @xmath64 in this case the first country has two monopolies ; country 3 one monopoly and countries 2 and 4 no monopolies at all .
all countries tend to zero exponentially faster , but the first one .
this is a simple example of an inward bellied matrix , that is , a matrix such that the diagonal crosses the external empty part .
because in this case the crossing country is the first one , all the other countries will converge to zero exponentially , as stated in the ansatz [ ansatz : cutpoint ] .
this phenomenon will be investigated in detail the next subsection .
+ in sec.[sec : ansatz ] we suggested a link between the convergence properties of countries and the shape of the belly of the ordered matrix . in this section
we verify the given ansatz by means of simple matrices , whose behavior is nevertheless analogous to the one we will find in real world applications . let us start with a set of 5x5 square matrices .
we point out that we use square matrices only because the diagonal and , as a consequence , the shape of the belly is evident by eye .
as we will see in the following , these results can be applied also to rectangular matrices by means of the generalized definition of diagonal discussed above .
+ @xmath65 matrix @xmath66 represents a borderline case , in which the same lines of reasoning adopted in sec.[2blocks ] can be applied . all countries but
one are converging to zero , preserving a well defined ranking .
+ matrix @xmath67 shows a clear inward belly .
this means that , according to our ansatz , in order to know which countries will converge to a nonzero value of fitness we have to eliminate countries starting from the ones with a lowest ranking until we find a submatrix with an outward belly . in this case
this is not possible , in the sense that the only fully convergent submatrix would be the trivial 1x1 matrix containing only country 1 and product 5 . as a consequence ,
only the country 1 ( which , in this case , is the trivial crossing country ) and product 5 converge to a nonzero value of fitness and complexity , while the other countries and products decay exponentially .
+ on the contrary , matrix @xmath68 has a clear outward belly : all countries have products beyond the diagonal and so converge to nonzero values of fitness .
an interesting situation is shown in matrix @xmath69 , which is equal to @xmath68 but for one product . by removing product 4 from country 2
is it possible to make all countries but one converge to a zero value of fitness .
in fact , in this case the diagonal crosses the external part of the matrix in correspondence with an high ranking country .
this situation highlights the importance , in general , of a careful data sanitation when empirical @xmath2 matrices are built , not only with respect to low fitness countries , but also for high fitness ones . in any case
, we point out that when one passes from matrix @xmath68 to @xmath69 the ranking is preserved . + in order to show a case in which the @xmath2 matrices are rectangular and the crossing country is nor in the top , neither in the lowest position in the matrix , but somewhere in the middle of it , we consider the three following matrices : @xmath70 while matrix @xmath71 shows full convergence of its countries , it is enough to remove one product to make the external area large enough to be crossed by the diagonal and , as a consequence , to make two countries converge to zero ( matrix @xmath0 ) .
the removal of one more product turns the convergence to zero exponential with the number of iterations .
we stress that , even if the number and the category of products of country 4 is the same for all the three situations , its value of fitness changes
. however , this fact is not paradoxical as long as the values of the matrices are calculated in a correlated way , by considering , for example , the revealed comparative advantage .
in this section we study how this variety of convergence behaviors affects real @xmath2 matrices .
typical @xmath2 matrices are bigger than the ones seen in chapter [ sec : numerical ] and present specific characteristics of nestedness and density ; nevertheless , our convergence ansatz turn out to be relevant for real @xmath2 matrices .
the first database we consider is based on the import - export flows of products among countries collected by the united nations and processed by baci @xcite and covers the years from 1995 to 2010 .
the number of products is fixed through the years and equal to 1131 , classified accordingly to the 4 digits harmonized system 2007 , while the number of countries slighlty varies between 146 and 148 .
a data sanitation procedure has been performed , and the resulting values of the @xmath2 matrices are assigned by means of a threshold on the revealed comparative advantage ( rca ) as introduced by balassa @xcite . in all the years
the matrix is outward bellied , accordingly to the definition implicit in [ ansatz : belly ] . in other words ,
the diagonal does not cross the external ( i.e. , empty ) part of the matrix , and so the algorithm converges for most countries to a non zero value of fitness . while the general behavior is clear , some exceptions are there for the least fit countries , as it is visible in figure [ fig : hs2007 ] . in the figure
we show a pictorial representation of the ordered @xmath2 matrix for the year 1995 .
the yellow ( red ) dots mean that the country corresponding to the given row is ( not ) exporting ( in the rca sense ) the product corresponding to the given column , that is , the matrix element is equal to 1 ( or zero , respectively ) .
the horizontal line corresponds to the _ crossing country _ defined in [ sec : ansatz ] , and the vertical one to the product with higher complexity exported by it .
the part of the matrix which converges to nonzero values of fitness and complexity is defined by these lines ( in particular , the top right portion of the original matrix ) .
the dashed line is the diagonal of the fully convergent matrix . obviously , after the removal process the matrix is outward bellied in all its parts .
matrix for the baci dataset , year 1995 .
the diagonal is consistently above the external part .
however , there is an exception for the low fit countries , zoomed in the picture inset .
the horizontal and vertical lines represent the number of countries and products that converge to zero.[fig : hs2007 ] ] it is interesting to note that the number of countries and products that need to be removed to have a matrix converging to non - zero values is reduced for the later years .
this behavior is general , for all the datasets checked .
the causes of this phenomenon will be discussed in the conclusions in section [ sec : conclusions ] .
a second dataset we analyze is the one collected by the united nations and processed by feenstra et al .
it contains the import - export flows of 538 products , classified accordingly the sitc rev2 products classification , for the years 1963 - 2000 .
the number of countries ranges from 131 to 154 .
the dataset is particularly interesting because of the long time span of data presented in a consistent form . in this dataset
the convergence ansatz proposed is more relevant , since big parts of it is consistently above the diagonal .
an example , that highlight also an application of ansatz [ ansatz : cutpoint ] , is presented in figure [ fig : sitc2 ] .
the example also highlights a peculiar graphical feature of the ordered matrix .
indeed , for the countries and products that are converging to zero it is possible to observe how the frontier between the internal and external part is smooth and continuous , while a more rough behavior can be observed for the areas converging to positive numbers .
this behavior is due to the fact that , since the fitnesses and complexities of the countries and products below the crossing country are converging exponentially to zero , for any given finite iteration the ratio between two consecutive fitnesses and complexities is also exponentially growing .
therefore the fitnesses are mostly determined by the highest complexity product and , similarly , the complexities are given by the lowest fitness country . as a consequence there is a clear correspondence between best products and worst countries , determining the smooth frontier that it is clearly visible in figure [ fig : sitc2 ] . at the opposite , for the countries above the crossing country , their relative fitnesses
converges to fixed ratios .
therefore the fitnesses of the countries are determined by all the products not converging to zero , causing their ranking to be not uniquely determined by their best product .
a similar line of reasoning can be used for the products .
therefore after the crossing country there is not a clear frontier .
a similar behavior can be observed in all the matrices that we studied . even in the case of figure
[ fig : hs2007 ] it is possible to recognize the phenomenon for the last four countries ( zoomed in the inset ) .
matrix for the 1965 - 2000 dataset , year 1965 .
the matrix frontier is consistently above the diagonal line .
the horizontal and vertical lines represent the number of countries and products that converges to zero .
the dashed line shows the diagonal of the new block remaining after removing those countries and products : the diagonal of the new @xmath2 matrix now does not cross the external part of the matrix.[fig : sitc2 ] ] the convergence to zero of a sizable set of countries is very evident for matrices particularly empty , like the one produce from the patents dataset .
it is a dataset organized by oecd starting from data produced by the european patent office joining countries and technological sectors if a firm in a country has patents granted in that sector .
it is then manipulated in the usual way , through consideration of comparative advantage to remove the size of the countries from the argument , generating a matrix that has a @xmath11 or @xmath12 if it produce more patents than expected in a particular technological sector .
since to patent a discovery in a sector the country has to be on the frontier of the technological progress in that sector , the matrix comes out with an inward belly , as it is visible in figure [ fig : patents ] .
matrix for the patents dataset , year 2002 .
the external part is very big , so diagonal line .
crosses it and a consistent fraction of countries converges to zero .
the dashed line shows the diagonal of the new block remaining after removing those countries and products.[fig : patents ] ] as for guess [ ansatz : cutpoint ] , a possible situation leading to a global convergence to zero is for a single country being diversified in too many sectors in which no other countries is interested in . in this case
we can have all the other countries fitnesses converging to zero .
an example is in figure [ fig : patents2 ] .
matrix for the patents dataset , year 2000 .
the diagonal line is consistently inside the external area .
more importantly , this is true even for just the top two countries , as it is visible in the zoomed in the picture inset . as a consequence ,
all the countries and products converges to zero but the first one , as shown by the horizontal and vertical lines .
the dashed line shows the diagonal of the block that would be left removing all countries and products but the first two : the diagonal would still cross the external part .
therefore the only remaining non - zero - fitness country is the first one , and the only remaining non - zero - complexity products are the ones produced only by it .
[ fig : patents2 ] ]
a practical application of our results is that the usual convergence conditions are not suitable for this algorithm . in particular
, the condition @xmath72 does not imply that the single components of the fitness vector stop to decrease and , in general , the resulting fitnesses and complexities could depend on @xmath73
. this could be particularly relevant if one is interested in non linear expressions of fitness or complexities ( e.g. the logarithm of the countries fitness ) , also given the fact that convergence rates can be very slow . in principle however , one may be interested only in the rankings . obviously , different criterions can be assessed for the ranking convergence .
indeed the problem is not trivial because of the very slow convergence of the fitness described in section [ sec : shape ] .
while the exact power law behavior @xmath74 described in section [ 2blocks ] is less likely to be found in larger real matrices , the exponential behavior @xmath75 can have a very large characteristic time @xmath76 , as noted in section [ sec : shape ] , so that can be considered approximately power law for a large number of iterations .
therefore the ranking can change even after many iterations of the algorithm .
it is better to see the issue in a practical case . in figure
[ fig : crossing_n ] , we show how the ranking keeps changing in a real case even after thousands of iteration steps .
indeed we can see how , in the context of a power law approximation , the approximated exponent @xmath77 can be very small for many iterations .
note that a value of @xmath77 of @xmath78 means that one must wait @xmath79 iterations for the fitness to reduce of one order of magnitude .
it is therefore clear that when two countries fitness are decaying with very similar @xmath77 , to switch ranks could take any amount of time .
we can see an example for two such countries , related to the previous real case , in figure [ fig : fitness_alpha_evolution ] . while it is impossible to assure that there will be no crossing in the next iterations , one can estimate a lower bound to the number of iterations still needed for the next change in ranking .
this is computed as follows : * we estimate the growth rates @xmath77 for any country @xmath8 , by the formula : @xmath80 where @xmath81 is a suitable delay . ] . *
we order the countries accordingly to their fitness to have @xmath82 . * for any @xmath8 we estimate the crossing iteration @xmath83 between the countries @xmath8 and @xmath84 accordingly to @xmath85 of course the quantity @xmath86 is a reasonable estimate of an actual possibility of a rank change only if @xmath87 , i.e. @xmath88 . in the other case , when @xmath89 , the two fitnesses are indeed diverging . * among the valid crossing iterations @xmath87 , we define the minimum crossing iteration as @xmath90 the value of @xmath91 is therefore , for any @xmath20 , a lower bound for the iteration when the next rank change will occour . its value can be used in order to asses when to stop the algorithm .
this value is indeed a lower bound since the real exponential decay is always faster than the estimated power law .
however , since the characteristic time of the exponential is very slow , the lower bound is very close to the real value .
this can be seen for example in the real case behavior of @xmath91 shown in figure [ fig : mci_evolution ] .
comparing figure [ fig : mci_evolution ] and figure [ fig : crossing_n ] gives an insight on the algorithm .
in particular one can precisely match the crossing times and the predicted minimum crossing times .
after each crossing @xmath91 jumps to predict the next crossing .
notice that the observed plateaux of @xmath91 show the precision of the power law approximation : the forecast would not change at all until the next crossing only if the power law behavior was exact . for a comparison ) .
the increasing behavior confirms the lower bound property of the measure . after the last crossing at the iteration 1892 , the next predicted change in ranking
is @xmath92 iterations . as a consequence ,
if i am interested only in the country ranking and i am not willing to run the algorithm for millions of iterations , it is reasonable to stop the algorithm here.,scaledwidth=70.0% ] as a practical note , if one is interested only in the ranking , one can set a threshold @xmath93 to the maximum number of iterations that he is a priori willing to take .
of course it is also possible to define an equivalent measure for the ranking of products , that would give a different stopping time for the algorithm . in the practical spirit of this section ,
it is important to stress that the number of iterations could get very high to achieve a satisfying value of @xmath93 . in this case
it is possible that , while some countries ( or products ) are still slowly decaying , others are already well below the precision of the machine used for computations .
some technical attention is therefore required to avoid that the algorithm fails to compute , since the fitness - complexity algorithm [ eq : fitness]-[eq : complexitynorm ] , is ill defined for zero values of fitness .
this paper shows that the intrinsic non linear structure of the algorithm proposed by @xcite has highly non trivial consequences on its convergence properties .
in particular , we have linked the shape of the ordered matrix on which the algorithm is based to the possible presence of countries whose fitness tends to zero .
obviously , also all the complexities of their exported products will converge to zero . in synthetic but general cases it is possible to show analytically that the presence of a large empty part of the matrix makes some countries converge to zero .
in particular , the condition to check is if the matrix diagonal crosses or not the empty part of the matrix .
we study numerically some simple cases which confirm our results also when the analytical approach was not possible and in some real world cases , such as two different country - product databases and a country - patent database .
we extend our results to generalized versions of the algorithm , finding similar results .
we also give a practical recipe to decide when to stop the algorithm if interested in a stable ranking .
a striking result of our analysis is that a large variety of situations is present .
for example , if one considers the un comtrade dataset 1995 - 2010 , almost all countries have a finite fitness , while in many years of the un comtrade dataset 1963 - 2000 more than a half of the countries converges to a zero value of fitness .
on the other hand , in some patents datasets all the countries fitnesses but one tend to zero .
tacchella , a. , cristelli , m. , caldarelli , g. , gabrielli , a. , and pietronero , l. ( 2012 ) .
a new metrics for countries fitness and products complexity . ,
( 2):723 .
feenstra , r.c . ,
lipsey , r.e .
, deng , h. , ma , a.c . , and mo , h. ( 2005 ) .
world trade flows : 1962 - 2000 .
technical report , national bureau of economic research .
gaulier , g. , and zignago , s. , ( 2010 ) .
baci : international trade database at the product - level .
centre detudes prospectives et dinformations internationales .
balassa , b. ( 1965 ) trade liberalisation and `` revealed '' comparative advantage .
the manchester school 33 : 99123 .
cristelli , m. , gabrielli , a. , tacchella , a. , caldarelli , g. , and pietronero , l. ( 2013 ) . measuring the intangibles : a metrics for the economic complexity of countries and products .
, ( 8):e70726 .
barro , r. j. and sala - i martin , x. ( 2004 ) . .
mit press .
hidalgo , c. and hausmann , r. ( 2009 ) .
the building blocks of economic complexity . , 106:1057010575 .
ricardo , d. ( 1817 ) . .
john murray .
many possible variations of the algorithm could be implemented changing equations [ eq : fitness]-[eq : complexitynorm ] with the purpose of not having zero solutions .
the most obvious one is probably a variation of the harmonic mean in eq [ eq : complexity ] with a different elasticity coefficient , such as @xmath94 for @xmath95 equation [ eq : complexityvariation ] goes back in equation [ eq : complexity ] , as in the original algorithm . for @xmath96 equation
[ eq : complexityvariation ] becomes a sum , symmetrically to equation [ eq : fitness ] .
this is in strong opposition to what the algorithm is meant to describe , since it would mean that the more countries produce a product the higher its complexity .
therefore we will limit ourselves to discuss the cases @xmath97 .
computations similar to the ones in section [ sec : theoretical ] can be done in this case , obtaining similar conditions .
equation [ eq : iteration ] is still valid , even if only in the limit @xmath98 , and therefore after many iterations .
is exactly valid for any @xmath99 , and not only in the limit @xmath98 , for the case in which the density of block 2 is @xmath12 . in that case , changing [ eq : a2 ] with [ eq : a2mod ] , everything works exactly as in the section [ sec : density_zeros ] ] however the parameter @xmath36 in this case is equal to @xmath100 also ansatz [ ansatz : belly ] and [ ansatz : cutpoint ] change accordingly . while @xmath39 with @xmath36 defined in equation [ eq : a2 ] identified the diagonal line , when we define @xmath36 through equation [ eq : a2mod ] we have that imposing @xmath39 we identify the curve @xmath101 examples for different values of @xmath102 are presented in figure [ fig : gammas ] . from the figure it is clear the trade - off of changing the value of @xmath102 to avoid values of the fitness converging to @xmath12 : to loosen the condition for high fitness countries
, one makes it stricter for low fitness countries , and viceversa .
moreover it is also clear that changing @xmath102 does not help if the frontier is above the diagonal in the middle . for different values of @xmath102 , for a matrix @xmath2 @xmath103 .
the algorithm described in equations [ eq : fitness]-[eq : complexityvariation ] converges to all values different from @xmath12 if the frontier is all below the line.[fig : gammas],scaledwidth=70.0% ] | we investigate the convergence properties of an algorithm which has been recently proposed to measure the competitiveness of countries and the quality of their exported products .
these quantities are called respectively fitness @xmath0 and complexity @xmath1 .
the algorithm was originally based on the adjacency matrix @xmath2 of the bipartite network connecting countries with the products they export , but can be applied to any bipartite network .
the structure of the adjacency matrix turns to be essential to determine which countries and products converge to non zero values of @xmath0 and @xmath1 .
also the speed of convergence to zero depends on the matrix structure .
a major role is played by the shape of the ordered matrix and , in particular , only those matrices whose diagonal does not cross the empty part are guaranteed to have non zero values as outputs when the algorithm reaches the fixed point .
we prove this result analytically for simplified structures of the matrix , and numerically for real cases .
finally , we propose some practical indications to take into account our results when the algorithm is applied . |
the presence of impurities in a crystal can influence its mechanical , magnetic , semiconducting , superconducting , dielectric , spectroscopic and transport properties . to cite just a few specific examples
, impurities can improve the voltage holdoff capability of insulating materials@xcite , they are critical for the optical properties of most laser materials and gemstones , and they turned out to be the secret of ancient damascus steel blades.@xcite still , the physics of impurity systems is not well developed .
the local crystallographic structure around impurities is unknown in most cases , although it is an essential piece of information to undestand their influence on the physical properties of the host and to carry out _ ab initio _ simulations of these materials .
this situation is due to the experimental and theoretical difficulties that are met to obtain microscopic information on impurity systems . from the theoretical point of view , quantum calculations of impurity systems require the use of large supercells that have long not been computationally manageable . from the experimental point of view , impurity systems can only be measured with methods which can selectively probe certain atomic species . among such methods ,
many of them , such as electronic paramagnetic resonance , give only indirect information on the position and nature of the atoms surrounding the impurity .
due to the advent of third - generation synchrotron radiation facilities , x - ray absorption spectroscopy ( xas ) is now able to investigate impurities in solids .
the extended x - ray absorption fine structure ( exafs ) region of a xas spectrum provides quantitative information about the short range organization around the absorbing atom ( coordination number , interatomic distances ) . the x - ray absorption near - edge structure ( xanes ) region usually gives qualitative information about the atomic arrangement up to the medium range order .
xanes is sensitive to the electronic structure ( bonding ) and probes the empty states of solids . in the case of dilute elements in single crystals , especially at the @xmath3 edge of 3@xmath4 transition elements , collecting exafs spectra with a good signal - to - noise ratio
is a difficult task , because the intensity of diffraction peaks becomes much larger than the exafs signal .
taking the example of impurities in corundum , @xcite good exafs spectra could be collected for 10000 wt.ppm cr@xmath5 impurities , but the usable energy range obtained for 1500 wt.ppm fe@xmath5 impurities was already quite limited , and the concentration of coloring impurities is often an order of magnitude lower than this . in such cases , xanes becomes a practicle and precious technique to get information about the local structure of the absorbing atom.@xcite however , the analyzis of the experimental xanes data is not straightforward because of the photoelectron multiple - scattering processes that occur in the near - edge region . _
ab initio _ xanes simulations are then required to relate the experimental spectral features to the local geometry around the aborbing atom .
recently , an efficient first - principle approach based on plane - wave pseudopotential formalism has been developed to calculate @xmath3-edge xanes spectra.@xcite here we apply this method to the @xmath3-edge of substitutional paramagnetic impurities in aluminum oxide .
more precisely , we investigate the structural modifications of the corundum crystal structure induced by the presence of substitutional ti@xmath5 , cr@xmath5 or fe@xmath5 . this system is chosen because corundum containing transition metal impurities is important for laser applications and as gemstones .
colorless corundum ( ) becomes red ruby , pink sapphire , or yellow sapphire when a small amount of cr@xmath5 , ti@xmath5 , or fe@xmath5 ions substitute for al@xmath5 ions , respectively.@xcite the color of blue sapphire is due to the presence of ( fe - ti ) pairs in . for this investigation ,
polarized xanes spectra are measured and calculated .
corundum belongs to the trigonal bravais lattice , the optical axis being parallel to the [ 111 ] direction of the trigonal unit cell ( i.e. to the @xmath6-axis of the hexagonal cell ) .
corundum is then a dichroic compound in the electric dipole approximation .
the dichroic signal is a direct signature of the departure of the impurity site from octahedral symmetry.@xcite xanes calculations are performed from structural models resulting from _ ab initio _ energy minimization calculations .
the agreement of our experimental and theoretical spectra demonstrates that precise structural information can be extracted from the anglular - dependent xanes spectra , provided the electronic potential is accurately modelled .
corundum ( ) belongs to the @xmath7 ( @xmath8 ) space group.@xcite the trigonal unit cell contains two al@xmath1o@xmath2 formula units .
the atomic environment of aluminum is illustrated in fig .
[ corund ] .
the atomic site of aluminum is a distorded octahedron with a 3 ( @xmath9 ) local point symmetry .
the alo@xmath10 octahedron is characterized by two different al - o interatomic distances . if we consider that the absorbing atom is denoted by al in fig .
[ corund ] , the nearest three oxygen atoms are labeled o@xmath11 in fig . [ corund ] .
the farther three oxygen atoms , labeled o@xmath1 , form a face shared by two octahedra along the three - fold symmetry axis . beyond the coordination shell of aluminum , the next two neighbors are aluminum atoms , labeled al@xmath11 and al@xmath1 in fig .
[ corund ] , and are relative to face - shared octahedra and edge - shared octahedra , respectively . aluminum atomic site in corundum .
left : the corundum structure is represented as a stacking of alo@xmath10 octahedra ( the trigonal cell and the three - fold symmetry axis are indicated ) .
right : zoom of the black - framed region of the left panel .
the aluminum site with its o@xmath11 , o@xmath1 , al@xmath11 and al@xmath1 first four neighbors are indicated . ]
three synthetic single crystals of doped ( obtained by the verneuil process ) are used for this study : a red ruby ( ) , a blue sapphire ( ) and a pink sapphire ( ) . the impurity concentration in each sample is given in table [ tab : compo ] .
they are measured with the cameca microbeam electron microprobe at the camparis analytical facility of paris .
a 30 kv acceleration with a 15 na beam current , defocused to 10@xmath12 m , is used .
x - ray intensities are corrected for dead - time , background , and matrix effects using the cameca zaf routine .
the standards used are , , and . the blue sapphire sample is chosen to represent the model compound of iron impurity in corundum ( : fe@xmath5 ) .
it is indeed preferred to a yellow sapphire sample , that only contains iron impurities , because synthetic yellow sapphires are known to be inhomogeneous.@xcite consequently , in this study , we assume that the low concentration of titanium ( 750 wt.ppm ) compared to that of iron ( 1500 wt.ppm ) does not affect the site relaxation around iron .
.[tab : compo]impurity concentration ( in wt.ppm ) in the three doped samples . [ cols="<,^,^,^",options="header " , ] first , the impurity site symmetry is conserved : the ti , cr and fe atoms are still in a @xmath9 point symmetry site , characterized by two kinds of m - o@xmath11 and m - o@xmath1 distances ( m = ti , cr or fe ) .
the angular relaxation is found to be very small ( i.e. less than 0.8 degrees ) .
this is illustrated in fig .
[ fig : angles ] , which gives @xmath13 as a function of the distance between the impurity and its neigboring atoms in the _ relaxed _ clusters .
second , a displacement of the impurity from the initial aluminum site is observed in the three _ relaxed _ models .
this displacement occurs along the @xmath9 axis , but its absolute value depends on the impurity .
chromium is displaced by 0.03 towards al@xmath11 atom , titanium is displaced by again 0.03 but in the opposite direction , and iron is further moved ( 0.09 towards al@xmath11 atom ) .
the small displacement of chromium is in agreement with various experimental studies of the literature.@xcite the result for iron is consistent with an analysis of electronic paramagnetic resonance ( epr ) experiments , that gives a displacement of @xmath14 .@xcite radial relaxation around the paramagnetic impurities in doped : norm of atomic displacement vectors ( oxygen atoms in solid line with circles and aluminum atoms in dashed line with triangles ) as a function of distance from the impurity in the _ relaxed _ model . ]
third , an increase of distances within the coordination shell of the impurity is observed .
this result is expected since the ionic radii of ti , cr and fe in octahedral site@xcite ( @xmath15 , @xmath16 and @xmath17 ) are all greater than the ionic radius of aluminum ( @xmath18 ) .
the calculated m - o@xmath11 and m - o@xmath1 bond lengths in doped were found to be close to m - o@xmath11 and m - o@xmath1 bond lengths in the corresponding metal oxides , @xmath0-m@xmath1o@xmath2 ( , and ) . on the contrary , the distances m - al@xmath11 and m - al@xmath1 were less affected by the relaxation process : they are closer to al - al@xmath11 and al - al@xmath1 bond lengths in corundum than the m - m@xmath11 and m - m@xmath1 distances in the corresponding oxides @xmath0-m@xmath1o@xmath2 .
tab.[distances ] gathers the m - o@xmath11 , m - o@xmath1 , m - al@xmath11 and m - al@xmath1 distances in : m@xmath5 deduced from the calculation , and compares them with the corresponding distances in , , and , resulting from x - ray diffraction experiments .
one can remark in tab .
[ distances ] that the ti - o@xmath11 and ti - o@xmath1 bond lengths in ( 1.97 and 2.07 ) are quite similar to those in ( 2.01 and 2.07 ) . besides , the ti - al@xmath11 and ti - al@xmath1 distances in ( 2.72 and 2.83 ) are closer to the al - al@xmath11 and al - al@xmath1 distances in corundum ( 2.65 and 2.79 ) than to the ti - ti@xmath11 and ti - ti@xmath1 distances in ( 2.59 and 2.99 ) .
one can notice that the experimental distances cr - cr@xmath11 in and al - al@xmath11 in are both equal to 2.64 .
thus one expects that cr - al@xmath11 distance in ruby ( ) should be around 2.64 .
the calculated distances of are slightly larger ( + 0.015 ) than those determined from exafs measurements on powder samples,@xcite and in good agreement ( + 0.005 ) with exafs measurements performed and on the same single crystal as the one used in the present study.@xcite they are also consistent with data extracted from optical spectroscopy.@xcite for : fe@xmath5 , the calculated distances are in agreement with exafs experiments@xcite and with epr results.@xcite finally , the m - o and m - al distances deduced from the _ ab initio _ relaxation calculations suggest that the structural relaxation around the transition metal impurities is quite local and mainly concern the coordination shell .
this point is confirmed by the calculations of the atomic displacements between the _ relaxed _ and _ non - relaxed_. fig . [
fig : dist ] represents the norm @xmath19 as a function of the distance between the impurity and its neighbors in the _ relaxed _ model .
it clearly shows that the oxygen coordination shell absorbs almost the relaxation .
the displacement of atoms farther than 2.5 from the impurity is less than 0.035 . comparing the three compounds
, one also observes that the relaxation is slightly larger around titanium and iron than around chromium .
the influence of the core - hole can be described by calculating a self - consistent potential around a core - hole or by solving a bethe - salpeter equation.@xcite for impurity systems , solving a bethe - salpeter equation is probably beyond the reach of available computer programs .
moreover , it was already observed that both methods give quite similar results.@xcite therefore , we shall describe the influence of the core - hole by calculating a self - consistent potential in the presence of a 1s core - hole .
calculated spectra performed with or without the presence of a @xmath20 core - hole on the chromium absorbing atom are compared in fig .
it clearly appears that the core - hole is needed to obtain a good agreement with the experimental curve . in particular , the intensity and the energy positions of the features within the 6005 - 6017 ev energy range are not well reproduced if the core - hole is not taken into account in the calculation .
we have noticed similar behaviors due to the presence of the core - hole at the titanium and iron @xmath3-edges in and : fe@xmath5 , respectively . on the other hand ,
the presence of the core - hole has a non - negligible impact in the pre - edge region .
this is illustrated in the inset of fig .
[ trou ] , which represents the pre - edge region ( 5985 - 5998 ev ) of the cr @xmath3 edge in .
an interpretation of this region in terms of group theory has been detailed in ref .
@xcite in the pre - edge region both electric dipole ( @xmath21 ) and quadrupole ( @xmath22 ) transitions occur .
since the site of the impurity in doped is not centrosymmetric , the @xmath23 states are present in the pre - edge through @xmath24 hybridization .
influence of the core - hole effects in xanes calculation at the cr @xmath3 edge in ruby .
inset : zoom of the pre - edge region . for pre - edge simulations ,
the calculated spectra include both electric dipole and electric quadrupole transitions ( the wave vector @xmath25 is along the [ 10@xmath26 direction of the trigonal cell of ruby@xcite ) .
] electric dipole ( @xmath21 ) and quadrupole ( @xmath22 ) transitions occur .
since the site of the impurity in doped is not centrosymmetric , the @xmath23 states are present in the pre - edge through @xmath24 hybridization .
the calculated curves shown in the inset of fig . [ trou ] are the sum of both contributions .
for the calculation of electric quadrupole transitions , the direction of the wave vector of the photon beam , @xmath25 , has to be specified . in the experiments ,
the single crystals were placed on the rotating holder so that the wave vector was parallel to the [ 10@xmath26 crystallographic direction .
consequently , the spectra @xmath27 shown in fig
. [ trou ] are calculated with @xmath28$].@xcite the two features of the experimental pre - edge are reproduced with or without the 1@xmath29 core - hole ( see inset of fig .
[ trou ] ) .
the presence of the core - hole essentially provokes a 2.5 ev shift towards lower energy of the pre - edge features , while the main rising edge is not shifted . a similar energy shift due to the core - hole
was already observed for electric quadrupole transitions at the ti @xmath3 pre - edge in rutile.@xcite unfortunately , here the shift induced by the presence of the core - hole is not large enough to give a perfect agreement with experiment .
this problem may be due to the limit of the density functional theory within lda to model the core hole - electron interaction for empty @xmath4 states of 3@xmath4 transition metals .
lda+u calculations could possibly improve the agreement in the pre - edge region . indeed in ref.@xcite , the authors show that lda+u calculations performed on ruby essentially affect the positions of cr @xmath4 level relative to the valence and the conduction band .
[ exp_calc ] compares experimental xanes data with the theoretical spectra obtained with the _ relaxed _ models .
the @xmath27 and @xmath30 signals are displayed at the ti @xmath3 edge for the pink sapphire sample ( fig.[exp_calc]a ) , at the cr @xmath3 edge for the ruby sample ( fig.[exp_calc]b ) and at the fe @xmath3 edge for the blue sapphire sample ( fig.[exp_calc]c ) .
one observes that the experimental spectra present comparable overall shapes .
the fe @xmath3 edge seems to be less resolved than the ti and cr @xmath3 edges , probably due to the following two reasons : ( i ) the core - hole lifetime is shorter at the fe @xmath3-edge ( 1.25 ev ) than at the cr @xmath3-edge ( 1.08 ev ) and at the ti @xmath3-edge ( 0.94 ev);@xcite ( ii ) the instrumental resolution is estimated to 0.9 ev at the fe @xmath3-edge while it is 0.6 - 0.7 ev at both the cr and ti @xmath3-edges .
we also noticed similarities between the m @xmath3 edges ( m = ti , cr , fe ) in : m@xmath5 and in @xmath0-m@xmath1o@xmath2 ( see ref.@xcite for the cr @xmath3 edge in and the fe @xmath3 edge in , and ref.@xcite for the ti @xmath3 edge in ) . at least to a certain extent , these spectra could be considered as the signature of a transition metal ion with a + iii oxydation state in six - fold coordinated environment .
however it is here necessary to perform calculations to go farther in the xanes analysis .
it can be seen in fig .
[ exp_calc ] that a good agreement is obtained between experimental data and calculated spectra carried out on the _ relaxed _ structural models .
the agreement is especially good for ruby at the cr @xmath3 edge , where energy positions and relative intensities of the various features are well reproduced .
it should be emphasized that not only the averaged @xmath27 spectra are correctly reproduced , but also the dichroic signal , from which more precise structural information could be extracted .
such a good agreement with angular dependent spectra has been also obtained with the same calculation method : at the al @xmath3 edge in corundum@xcite and at the si @xmath3 edge in @xmath0-quartz.@xcite comparison between @xmath31 ( left ) and @xmath32 ( right ) experimental data ( thick solid line ) and calculated spectra obtained with both _ relaxed _ model ( thin solid line ) and _ non - relaxed _ model ( dashed line ) : a ) the ti @xmath3 edge in pink sapphire , ; b ) the cr @xmath3 edge in ruby , ; c ) the fe @xmath3 edge in blue sapphire , .,title="fig : " ] comparison between @xmath31 ( left ) and @xmath32 ( right ) experimental data ( thick solid line ) and calculated spectra obtained with both _ relaxed _ model ( thin solid line ) and _ non - relaxed _ model ( dashed line ) : a ) the ti @xmath3 edge in pink sapphire , ; b ) the cr @xmath3 edge in ruby , ; c ) the fe @xmath3 edge in blue sapphire , .,title="fig : " ] comparison between @xmath31 ( left ) and @xmath32 ( right ) experimental data ( thick solid line ) and calculated spectra obtained with both _ relaxed _ model ( thin solid line ) and _ non - relaxed _ model ( dashed line ) : a ) the ti @xmath3 edge in pink sapphire , ; b ) the cr @xmath3 edge in ruby , ; c ) the fe @xmath3 edge in blue sapphire , .,title="fig : " ] we have seen in sec . [ subsec : res_models ] that the differences between the _ relaxed _ and _ non - relaxed _ structures are essentially concerned with the interatomic distances within the coordination shell of the impurity .
therefore , it is interesting to test the sensitivity of xanes to such subtle structural modifications . fig .
[ comp_models ] compares @xmath33 and @xmath32 experimental spectra with the corresponding calculated ones performed from both _ non - relaxed _ and _ relaxed _ models .
first of all , it is worth noting that the differences observed between calculated spectra are small but not negligible .
for all compounds , peaks labeled @xmath34 in @xmath31 and labeled @xmath35 and @xmath36 in @xmath32 are too strong in the _ non - relaxed _ calculated spectra .
consequently , these features ( all located at about 30 ev above the edge ) are clearly correlated to the first interatomic distances around the impurity .
the same conclusion can be made for the shoulders labeled @xmath37 and @xmath38 in @xmath32 and @xmath39 in @xmath31 . especially at the cr @xmath3 edge in @xmath32 signal
, one can see that the double - feature ( @xmath40 and @xmath41 ) is better reproduced by the _ relaxed _ calculation than by the _ non - relaxed _ one . in a previous study,@xcite
the _ relaxed _ model was validated up to the second neighbors around the impurity by comparison with exafs measurements . here ,
thanks to the very good agreement obtained between experimental and _ relaxed _ calculated xanes spectra , the _ relaxed _ structural model is again validated and this time to greater extent .
in this paper , we have shown that xanes can provide precious information about the small structural relaxation occuring within the atomic site of impurity in an aluminum oxide matrix .
this has been achieved by ( i ) measuring angular dependent xanes spectra in order to probe the distortion of the impurity site , ( ii ) calculating structural model by ab initio energy minimization , ( iii ) calculating the xanes spectra from the previous theoretical structural models using the full - potential pseudopotential self - consistent method of ref.@xcite the good agreement obtained between experimental and calculated spectra permits the validation of the structural models .
we have also pointed out the importance of the electronic potential construction to carry this study to a successful conclusion .
these results opens new applications in xanes analysis , in particular for dilute samples for which good quality exafs measurements are difficult to collect .
we wish to acknowledge the computational support of the french _ institut du dveloppement et de recherche en informatique scientifique _ in orsay , where all the calculations of this study were carried out . | we determine the structural relaxations around paramagnetic impurities ( ti , cr , fe ) in corundum ( @xmath0-al@xmath1o@xmath2 ) , by combining x - ray absorption near edge structure ( xanes ) experiments and _ ab initio _ calculations .
the structural relaxations are found to be very local .
we then show that xanes is sensitive to small variations in interatomic distances within the coordination shell of the absorbing atom .
the experiments were carried out on single crystals of ruby and sapphires .
linear dichroic signals are essential to characterize the geometry of the impurity site .
the calculations were performed within a self - consistent `` non muffin - tin '' framework , that uses pseudopotentials , plane - wave basis set , and the continued fraction for the absorption cross section . |
the qcd dirac operator contains all information about the quark dynamics in qcd
. therefore its spectrum is of high interest . in the literature
one considers the euclidean dirac operator which reads @xmath2 .
we will use euclidean space time throughout this work .
since the dirac operator is antihermitean , its eigenvalues are purely imaginary . with the eigenvalue equation @xmath3 one defines the spectral density @xmath4 where the averaging is over the gluon background .
the first discovery about the spectral density has been made by banks and casher @xcite who connect the order parameter of spontaneous chiral symmetry breaking , i.e. the chiral condensate , to the qcd spectrum at zero eigenvalues . since then , different approaches have led to a steady progress in this field . as the dynamics at low energies
is mainly driven by chiral symmetry the behaviour of the spectral density in the vicinity of @xmath5 can be analysed using chiral perturbation theory .
the first correction to the banks casher relation was found by smilga and stern @xcite @xmath6 where @xmath7 denotes the euclidean 4volume .
the constant term is the banks casher result , whereas the linear term vanishes in the case @xmath8 .
+ chiral random matrix theory considers the so called microscopic spectral density , in which all eigenvalues are rescaled according to the average spacing of small eigenvalues .
it successfully describes the dirac spectrum in the extreme infrared regime , where the spectral density is completely determined by the global symmetries of the dirac operator ( see e.g. @xcite ) .
universal properties of the microscopic spectral density have been derived and identified in lattice calculations @xcite .
furthermore partially quenched chiral perturbation theory combines the results of chiral perturbation theory and chiral random matrix theory @xcite .
+ the dirac spectrum for small eigenvalues is driven by chiral symmetry which determines hadron dynamics for small momenta .
thus chiral perturbation theory and chiral random matrix theory can be successful in describing the dirac spectrum .
+ the same holds for the nambu jona - lasinio ( njl ) model which serves as an effective model for chiral symmetry . in this work
we consider it in its bosonized form given by the linear @xmath0-model . the qcd dirac spectrum and the spectrum for the linear @xmath0model which we analyze in this paper should coincide in the regime , where the linear @xmath0model is a good effective description for qcd , that is for @xmath9 here @xmath10 gev defines the momentum scale below which a hadronic description of qcd may be useful .
the lagrangian of the linear @xmath0model in euclidean spacetime is @xmath11 q + \frac{1}{2}z_\phi \left ( ( \partial_\mu \sigma)^2+(\partial_\mu \vec\pi)^2 \right)+v(\sigma,\vec\pi , m_q ) \ ; .\ ] ] the fermion fields describe constituent quarks which appear in two flavours and three colours .
the bosonic potential @xmath12 contains a chiral symmetric part @xmath13 depending on the @xmath14-symmetric representation @xmath15 and a linear symmetry breaking term @xmath16 . for later purposes we explicitly stress the dependence of the meson potential @xmath17 on the current quark mass @xmath18 . at the ultraviolet scale , bosonization of the njl model
fixes the form of the meson potential @xmath19 to be @xmath20 the linear symmetry breaking term @xmath21 is related to the current quark mass @xmath18 by @xmath22 we use renormalization group ( rg ) flow equations for the linear @xmath0-model with quarks in the schwinger proper time formalism .
these flow equations have been discussed in various papers .
for a general review on rg flow equations see @xcite and for the schwinger proper time approach used here see @xcite .
the evolution equations provide a framework to study nonperturbative theories by integrating out smaller and smaller momentum shells above a limiting infrared cutoff @xmath23 .
this cutoff enters in the form of a smooth regulator function @xmath24 in the schwinger proper time representation of the logarithm containing one - loop fluctuations .
we consider the set of regulator functions given by @xmath25 the flow equations which we present in this paper are obtained with @xmath26 . in order to check the cutoff dependence of our results we use regulator functions with different n. generalization of the equations
is straight forward .
+ note , the parameter @xmath21 is scale independent , since a linear symmetry breaking term does not evolve under rg flow @xcite .
therefore only the chiral symmetric part @xmath27 of the meson potential @xmath28 evolves under flow of the ir cutoff scale @xmath23 .
+ in this paper we restrict ourselves to the flow equations in the large @xmath1 approximation , which have been considered recently @xcite and whose results have proven equivalent to the standard selfconsistent njl large @xmath1 approach @xcite .
they are particularly simple and read @xmath29 the couplings @xmath30 and @xmath31 in eq .
( [ eq : lagrangian ] ) do not evolve in large @xmath1 and have been set to 1 for convenience .
+ an additional flow equation which determines the chiral condensate can be obtained by differentiation of the partition function @xmath32 with respect to an appropriately introduced source term @xmath33 which probes the chiral condensate @xmath34 this flow equation is derived analogously to the other flow equations by introduction of the schwinger proper time cutoff in @xmath32 and reads @xmath35 in large @xmath1 all flow equations can be solved analytically .
the initial conditions and solutions are discussed in @xcite . denoting @xmath36 one finds for the potential ( at @xmath37 )
@xmath38 the minimum of this potential yields the vacuum expectation value @xmath39 of the bosonic field which is nonvanishing due to the spontaneous and explicit breaking of chiral symmetry .
the physical value for the quark condensate is obtained as @xmath40 .
its flow is shown in fig . 1 .
@xmath41 $ ] + we obtain an infrared value @xmath42 mev , which compares well to the value obtained from the gell - mann oakes renner relation @xmath43 mev .
this result is not surprising ( cf .
@xcite ) since the initial conditions of the flow equations have been adjusted to reproduce @xmath44 mev and @xmath45 mev at @xmath37 using a current quark mass @xmath46 mev at the ultraviolet scale .
while the qcd dirac operator @xmath47 is antihermitean , the dirac operator in the linear @xmath0model @xmath48 can have complex eigenvalues . in the flow equations we have neglected the imaginary part of the fermion determinant , which is related to anomalous processes @xcite . in this approximation
the eigenvalues of the dirac operator in the linear @xmath0-model are purely imaginary in coincidence with the qcd spectrum . for general complex eigenvalues
@xmath49 it is useful to consider the two dimensional spectral density @xcite @xmath50 which can be obtained from the complex resolvent @xcite @xmath51 \biggr \rangle_\phi \;.\end{aligned}\ ] ] the complex eigenvalue @xmath52 is the generator in this resolvent , which in our low energy effective theory involves an average over the meson fields denoted by @xmath53 . in large @xmath1 approximation
there are no meson fluctuations .
therefore the averaging over the meson fields can be interchanged with the logarithm in the last line and one finds @xmath54 the connection between @xmath55 and @xmath56 can be derived with eq .
( [ eq : complex - spectral - density ] ) @xmath57 combining eqs .
( [ eq : r2c ] ) and ( [ eq : g2p ] ) yields @xmath58 this relation can be inverted @xcite and gives @xmath59 where @xmath60 the complex eigenvalue @xmath52 enters in the resolvent eq .
( [ eq : resolvent ] ) just as an explicit quark mass .
therefore no new flow equations have to be evaluated .
after bosonization @xmath52 appears in a linear symmetry breaking term in the meson potential via eq .
( [ eq : mass2linear ] ) with @xmath61 . +
as the resulting meson potential @xmath62 becomes complex , we must comment on the meaning of @xmath63 . instead of restricting ourselves to positive real values @xmath64
, we now have to allow also complex fields @xmath63 . instead of finding a real minimum of the potential @xmath65 we have to determine a complex saddle point , which minimizes the real part of the complex meson potential @xmath62 .
+ from eq .
( [ eq : p ] ) it is obvious , that the two dimensional spectral density is zero if @xmath66 is analytic in @xmath52 . in presence of the complex quark mass @xmath52
the chiral condensate at @xmath37 becomes in the large @xmath1 limit with eq .
( [ eq : kondensat - flussgleichung ] ) @xmath67\;.\ ] ] this expression is analytic if @xmath68 is analytic and @xmath69 .
as @xmath68 is the bare pion decay constant in presence of @xmath52 and therefore minimizes @xmath70 , the second condition @xmath69 is always true .
but the global minimum of @xmath71 can switch if @xmath52 changes infinitesimally .
this is exactly what happens and leads to a discontinuity in @xmath68 .
fig [ fig : epsilon ] demonstrates this behaviour .
[ fig : epsilon ] @xmath0 @xmath72 + for @xmath73 it follows from eq .
( [ eq : potsol ] ) in the chiral limit that with @xmath39 as a global minimum of @xmath71 also @xmath74 , @xmath75 and @xmath76 are global minima of @xmath71 .
additionally they are saddle points of @xmath77 . for
either @xmath78 or @xmath79 the global minimum is still twice degenerate . in case of an infinitesimal @xmath80 the sign of
@xmath80 chooses the minimum , i.e. @xmath81 changes its sign under the transformation from @xmath82 to @xmath83 .
the same holds for infinitesimal @xmath84 and for @xmath85 , but in this case an infinitesimal @xmath84 only gives rise to an infinitesimal @xmath86 , consequently no discontinuity arises in @xmath87 .
we are thus left with the result , that @xmath56 can only be nonzero for @xmath88 .
the position of our dirac spectrum thus coincides with the qcd spectrum . using the ansatz @xmath89 we find from eqs .
( [ eq : p2qbarq ] ) , ( [ eq : p ] ) and ( [ eq : def_zstar ] ) @xmath90\;.\ ] ] now let us consider the flow equations for the potential and the chiral condensate . in the large @xmath1 limit they reveal a simple connection , namely . as the expectation value of the pion field vanishes
the notation above is unambiguous . ]
@xmath91 with the initial conditions @xmath92 and @xmath93 at @xmath94 this equation can be integrated out and yields @xmath95 eq .
( [ eq : spektrum ] ) gives the spectral density in terms of the discontinuity of the generalized chiral condensate @xmath96 across the real axis , which can be calculated from the saddle point @xmath68 of @xmath97 fulfilling @xmath98 . with eq .
( [ eq : gorcheck ] ) we find in the chiral limit @xmath99 \nonumber \\ & = \frac{m_\lambda^2v_4}{\pi}\text{re}\left ( \phi(i y+\epsilon ) \right)\ ; .\end{aligned}\ ] ] the result is plotted in fig .
( [ fig : spektraldichtegross ] ) .
[ fig : spektraldichtegross ] @xmath100 $ ] @xmath100 $ ] + next we derive the behaviour of the spectral density at small eigenvalues .
therefore we have to evaluate @xmath101 for small @xmath84 .
the value @xmath102 is determined by the saddle point equation @xmath103 @xmath104 and @xmath105 can be expanded in a taylor series .
the coefficients are obtained from @xmath106_{y=0}=0 \quad , \quad n=0,1,2 \ ; .\ ] ] we find @xmath107 ^ 3}y^2+ ... \ ] ] with eq .
( [ eq : rho ] ) the spectral density @xmath108+\frac{1}{2}\frac{m_\lambda^4 \left.\partial_\phi^3 u(\phi)\right|_{\phi=\phi_0(0)}}{\left [ \left.\partial_\phi^2 u(\phi)\right|_{\phi=\phi_0(0 ) } \right]^3}y^2+o(y^3 ) \right]\ ] ] can be calculated explicitly for small eigenvalues , see eq .
( [ eq : potsol ] ) and ( [ eq : gorcheck ] ) @xmath109 ^ 3 } \,y^2+o(y^3 ) \;.\nonumber \end{aligned}\ ] ] the constant coefficient is the banks casher result @xcite .
since our computation is done for @xmath8 there is no linear term in agreement with the findings of smilga and stern for @xmath8 @xcite .
numerically , the quadratic coefficient is 87 mev for the choice of parameters given in @xcite .
the quadratic term in @xmath84 can be related to the correlation function of three scalar @xmath0quanta . in the numerator of eq .
( [ eq : quadrat ] ) one sees the effective @xmath110 coupling constant contained in our potential @xmath111 multiplied by the cube of the @xmath0propagator @xmath112 at zero momenta .
this result can also be found in the paper of smilga and stern @xcite .
we have presented a method to obtain the spectral density of the dirac operator for the linear @xmath0-model in the large @xmath1 limit . because of asymptotic freedom in qcd
, we expect @xmath113 for asymptotically large eigenvalues , since for a free theory the spectral density is only determined by the phase space .
this situation of a non interacting theory can be tested within our approach in the limit @xmath114 . in this case
the njl four fermion interaction @xmath115 @xcite vanishes . as shown in fig .
( [ fig : free ] ) the resulting spectral density becomes the density of a free theory . for @xmath116
the phase space is cut off sharply .
@xmath100 $ ] [ fig : cutoffs ] @xmath100 $ ] + if the interaction strength @xmath117 is turned on by a finite @xmath118 one sees that quark - quark interactions lead to a level repulsion increasing the density for small @xmath119 and pushing some strength above 1 gev . varying @xmath120 and thereby the shape of the cutoff functions @xmath121 in eq .
( [ eq : cutoffs ] ) we adjust the cutoff parameters @xmath118 and @xmath122 in such a way that @xmath123 and @xmath124 are kept constant . this way we can test the sensitivity of the spectral density to the cutoff function .
we see that below 1 gev the choice of the cutoff function leads to stable results within @xmath125 .
+ in the realistic case of qcd we expect that gluonic interactions above @xmath122 influence the spectrum nearby and below . therefore , the decrease of the spectral density @xmath126 above @xmath127 gev is probably an artifact of the model space , which is resticted by the soft cutoff . + the interactions in the strong coupling regime of qcd induce level repulsion .
higher modes are strongly suppressed , while the increase of the average eigenvalue density for smaller eigenvalues drives the chiral condensate which appears on the low energy end of the spectrum .
it seems that level repulsion and condensation are the dominating effects for the dirac operator spectral density in the strong coupling regime of qcd .
in this paper we have presented a new method to obtain the dirac spectrum with renormalization group flow equations .
we have used proper time flow equations for the linear @xmath0-model with quarks of two flavors . in the large @xmath1 approximation
we have found a simple connection between the meson potential @xmath128 and the chiral condensate which allowed us to calculate the dirac spectrum directly from the full effective meson potential @xmath97 .
we have recovered the banks casher relation for small eigenvalues .
the next to leading order coefficient is 87 mev .
+ the calculation of @xmath126 should be valid up to several hundred mev .
thus our approach allows to enter a new regime in the description of the dirac spectrum which may complement other results from chiral perturbation theory and chiral random matrix theory .
compared to a free dirac theory we have found the eigenvalue density suppressed for large eigenvalues and increased for small eigenvalues .
all results are obtained in the infinite volume limit .
since the schwinger proper time cutoff functions are simple to handle , the same calculation can be done for finite euclidean space time volumes , where the evolution of the flow equations enters decisively .
such a calculation could then be compared to qcd lattice computations @xcite .
further we plan to extend the calculation beyond the large @xmath1 approximation .
the renormalization group flow equations of the linear @xmath0-model with quarks provide a framework to include meson loops in the njl - model without an additional new cutoff @xcite .
we would like to thank j. meyer , s. meyer , h.a .
weidenmueller , m.a .
stephanov , g. akeman and m.a .
nowak for stimulating discussions .
t. banks and a. casher , nucl .
b * 169 * , 103 ( 1980 ) .
a. v. smilga and j. stern , phys .
b * 318 * , 531 ( 1993 ) .
j. j. verbaarschot and t. wettig , ann .
nucl . part .
sci . * 50 * , 343 ( 2000 ) , hep - ph/0003017 .
m. e. berbenni - bitsch , s. meyer , a. schafer , j. j. verbaarschot and t. wettig , phys .
lett . * 80 * , 1146 ( 1998 ) hep - lat/9704018 .
h. damgaard , hep - lat/0110192 .
j. c. osborn , d. toublan and j. j. verbaarschot , nucl .
b * 540 * , 317 ( 1999 ) hep - th/9806110 .
j. berges , n. tetradis and c. wetterich , hep - ph/0005122 ; j. berges , lectures given at 11th summer school and symposium on nuclear physics ( nuss 98 ) , hep - ph/9902419 .
b. schaefer and h .- j .
pirner , nucl . phys .
a * 660 * ( 1999 ) 439 , nucl - th/9903003 ; b .- j .
schaefer and h .- j .
pirner , nucl .
a * 627 * ( 1997 ) 481 hep - ph/9706258 ; j. meyer , g. papp , h .- j .
pirner and t. kunihiro , phys .
c * 61 * ( 2000 ) 035202 , nucl - th/9908019 ; g. papp , b .- j .
schaefer , h .- j . pirner and j. wambach , phys .
d * 61 * , 096002 ( 2000 ) , hep - ph/9909246 .
j. zinn - justin , `` quantum field theory and critical phenomena '' , oxford 1989 .
j. meyer , k. schwenzer , h. j. pirner and a. deandrea , phys .
b * 526 * ( 2002 ) 79 , hep - ph/0110279 .
s. p. klevansky , rev .
phys . * 64 * 649 ( 1992 ) ; g. ripka , `` quarks bound by chiral fields '' , oxford university press ( 1997 ) . | we derive the spectrum of the dirac operator for the linear @xmath0model with quarks in the large @xmath1 approximation using renormalization group flow equations . for small eigenvalues ,
the banks casher relation and the vanishing linear term are recovered .
we calculate the coefficient of the next to leading term and investigate the spectrum beyond the low energy regime . /#10=
0=0 1= 1=1 0>1 # 1 / |
the experimental lower limit on the lifetime of the proton is a severe problem for models in which the fundamental scale of quantum gravity is low compared to the supersymmetric gut scale @xmath3gev @xcite .
a baryon number u@xmath4 symmetry can not be gauged in field theory ; like other accidental symmetries it is expected to be violated by effects at the string scale or by quantum wormholes and virtual black holes @xcite . such @xmath5 violation would appear at low energies as nonrenormalizable operators , for example of the form @xmath6 where @xmath7 is the fundamental scale and @xmath8 is expected to be @xmath9 in the absence of suppression mechanisms . then for @xmath7 in the @xmath10tev range , for which collider signals of the fundamental degrees of freedom or of large extra dimensions may be observable , @xmath11 comes out to be under a second , to be compared with the ( mode - dependent ) experimental bounds of order @xmath12 _ years _ @xcite .
various solutions have been proposed @xcite all of which have implications for the production of an excess of baryons over antibaryons in the early universe , for which @xmath5 violation is a precondition .
the production of gravitational kaluza - klein modes in the early universe in such models , in which some compactified extra dimensions are orders of magnitude larger than the fundamental length , gives a severe upper bound on the maximum temperature attained consistent with cosmological observations .
even for the maximum number ( usually considered as 6 ) of large extra dimensions and the relatively large value @xmath13tev , a maximum temperature of a few gev is the upper bound if overclosure of the universe , disruption of the successful predictions of nucleosynthesis , and an observationally unacceptable level of background gamma - rays from k - k mode decay are to be avoided @xcite . the energy density in k - k modes would also affect the evolution of density fluctuations by altering the time of matter - radiation equality @xcite .
astrophysical production and decay of such modes also leads to an independent lower bound on the fundamental scale @xcite , which is also constrained by the non - observation ( to date ) of direct and loop effects in experiments @xcite .
any attempt at explaining proton longevity and baryogenesis should operate within these constraints . exact ( anomaly - free )
discrete or horizontal symmetries can be imposed to forbid @xmath5-violating operators mediating proton decay @xcite while allowing others , through which baryogenesis occurs : this approach requires an `` x - boson '' to be present , with couplings which appear unnaturally small ( in contrast to the standard gut or leptogenesis scenarios ) .
baryon number can be gauged if the anomaly is canceled by a string theory mechanism , or @xmath5 violation may be forbidden to all orders in perturbation theory by string selection rules , in some `` intersecting brane '' models @xcite .
note however that in a more general class of intersecting brane models @xcite , such selection rules do not prevent the four - fermion operators from appearing , as discussed above , in which case the fundamental scale can not be low . if baryon number is perturbatively exact , nonperturbative processes @xcite are the only option to create net baryon number .
it is difficult to see how this proposal can be reconciled with cosmological constraints , since any such processes would operate at or above the electroweak scale and be enormously suppressed at low temperature .
* first reference ) actually satisfies the bounds , since a hot plasma at temperature @xmath14 is needed . ]
a novel mechanism for suppressing proton decay in the context of a low fundamental scale theory has been proposed in ref .
this is a geometrical mechanism for suppressing 4d @xmath5-violating operators , namely localization of fermions in extra dimensions .
the simplest implementation is for the su@xmath15 gauge fields to propagate in one extra dimension ( _ cf . _
@xcite ) , in which the quark and lepton wavefunctions are peaked about points separated by a distance @xmath16 ( fig .
[ fig : branesetup ] ) . then any strong @xmath5-violating operators in the effective 5d theory can only produce proton decay proportional to the overlap of the wavefunctions , which can be exponentially small .
alternatively , proton decay by exchange of massive modes is suppressed by the yukawa propagator over the distance @xmath17.-violating interactions propagate over the bulk . ]
nonperturbative quantum effects which may lead to proton decay , for example virtual black holes @xcite , are also exponentially suppressed due to the integration over the fifth dimension .
baryogenesis in such scenarios has been discussed in ref .
however , the authors of that work introduced a standard model singlet scalar with renormalizable baryon number - violating interactions and primarily considered the effect of thermal corrections to the wave function overlap which controls the strength of the baryon number violation . an affleck - dine - type scenario consistent with large extra dimensions , which involves introducing a scalar charged under a new @xmath18 group , has recently been described in @xcite ; while the idea appears viable from a cosmological viewpoint , the authors do not discuss how @xmath5-violating couplings which induce proton decay are to be suppressed , while still keeping the operators by which the scalar decays .
one might imagine that additional exact symmetries could be imposed to achieve this , in the spirit of @xcite .
in this work , we consider a related , but simpler class of scenarios and report mostly on a `` no go '' result .
specifically , we consider the feasibility of baryogenesis under the following conditions : 1 .
the only renormalizable operators that are present at the electroweak scale are those of the standard model , and the perturbative 4d effective field theory ( dimensionally reduced from an intrinsically higher - dimensional theory ) is valid .
( the possibility of doing without an `` x - boson '' was briefly mentioned in @xcite . ) in other words , we work with an effective theory with top as the heaviest field that has not been integrated out and neglect any lorentz violating operators resulting from time dependence of the extra dimensional geometry .
this condition is motivated from our ignorance about the physics beyond the standard model and the fact that no viable baryogenesis scenario has yet been proposed with only the standard model fields and gravity .
the leading baryon number - violating operator due to physics at the fundamental scale is a dimension-6 operator of the form @xmath19 where @xmath20 the wave function overlap , which is a time dependent function .
the wave function overlap @xmath20 can be time - dependent whenever fermion localization is controlled by a field , akin to the radion , which is initially displaced from the minimum of its effective potential .
the work of @xcite can also be taken to be of this type , where the `` radion '' is always at the minimum of a time dependent effective potential , which is derived from time dependent finite temperature effects .
note that this condition implies the assumption that 4d instanton / sphaleron induced baryon number violating operators are weaker than the dimension-6 operator , _
i.e. _ only perturbative physics is analyzed for dimension @xmath21 operators .
this is a reasonable assumption as long as the temperature is well below the electroweak scale ( which is anyway required in low fundamental scale models ) and the su@xmath22 coupling remains small ( appendix [ app : inst ] ) .
the 4d planck scale @xmath23 has an adiabatic time dependence due to the change in the volume of extra dimensions , and the fundamental scale of gravitational physics is at @xmath24gev .
the upper bound on this scale comes from the desire to ameliorate the hierarchy problem without susy , while the lower bound is obtained by the bounds on neutron - antineutron oscillations , which can not be suppressed by a geometrical mechanism .
note that there is still a certain amount of fine tuning , of order @xmath25 , with such a ( relatively ) large fundamental scale .
there is no electroweak phase transition because of a low reheating temperature ( _ i.e. _ @xmath26gev ) @xcite .
baryogenesis is also assumed to occur above the qcd phase transition temperature ( _ i.e. _ @xmath27gev ) . 5 .
the effective 3d space is homogeneous and isotropic at all times when the temperature of the universe is @xmath27gev . in this context
, we analyze the possibility of baryogenesis and find specific conditions under which it is feasible .
our setup contains all the necessary ingredients of baryogenesis @xcite : 1 .
baryon number violation ( dimension-6 operator ) 2 .
@xmath28 and @xmath29 violation ( standard model ckm + dimension-6 operator ) - the resulting quantity is much larger than the jarlskog invariant .
3 . nonequilibrium ( spacetime expansion or other unspecified means ) the reason for the difficulty of this scenario will be ingredient 1 , the baryon number violating operator being too weak in the low temperature setting , independently of ingredients 2 and 3 . since the usual jarlskog invariant of the sm is not the relevant quantity , given the presence of dimension-6 operators , the reason why this baryogenesis scenario is difficult has nothing to do with traditional wisdom . in particular , we find that dimension-6 operators are in most cases too weak to compete with baryon number - conserving operators which create entropy , for most sets of reactants .
navely , one finds this result a little surprising because one would expect @xmath30 can be easily engineered to be sufficiently large , where @xmath31 is the dimensionless quantity associated with @xmath0 violation , @xmath32 is the maximal branching ratio into the baryon number violating channel fixed by our model , @xmath33 is the number density of particles that can be converted into baryon number , @xmath34 is the number of degrees of freedom contributing to the entropy , and @xmath35 is the temperature at some fiducial time after which the baryon number is conserved .
however , even though @xmath31 can be naturally large in our scenario ( not governed by the jarlskog invariant ) , @xmath36 can not in general be made large because as the baryon number violating reactions are creating baryon asymmetry , the baryon number conserving counterparts are creating proportionately large amounts of entropy density . in other words , having the quark - lepton separation go to zero does not generate order 1 branching of the baryon number violating channels . instead , such channel is always suppressed by a scale much larger than the electroweak scale .
we do find an exception to our `` no go '' rule for a special situation in which the universe starts out dominated by the @xmath2 and @xmath1 quarks and gluons ( which might happen through the peculiarities of reheating after inflation ) , and the lighter quark @xmath1 is out of equilibrium while the heavier quark @xmath2 is in equilibrium , for example due to the anomalously large expansion rate and the anomalously small density of @xmath1 quarks .
in this exceptional scenario , there are two key ingredients compensating the suppression of the dimension 6 operator : 1 . due to the small @xmath37 ,
the @xmath2 quark interactions with out of equilibrium @xmath1 quarks do not generate significant entropy ; 2 .
the integration time for baryogenesis can be made long if the expansion rate is small , due to the nonstandard cosmology of extra dimensions .
however , this remote possibility , which requires additional assumptions about special initial conditions and/or more new physics , is unlikely to be realizable , and we do not have a detailed model to demonstrate its full viability . throughout this paper , we do not assume thermal equilibrium initial conditions , except where explicitly noted .
hence , although the short distance operators are fixed , the nonzero initial particle densities that exist in the universe can be _ a priori _ arbitrary combinations of species .
thus , when we consider bounds , we explicitly check all likely combinations of species without any bias to their initial number distributions , except when otherwise noted .
as far as the out - of - equilibrium - condition due to the expansion of the universe is concerned , we assume that somehow the planck scale can be adjusted to dial into the appropriate expansion rate .
of course , in a more completely specified model , this may turn out to be problematic , due to the bounds on moduli overproduction and other light moduli problems .
however , our results are robust with respect to how the moduli problems and initial conditions problems can be solved . in fact
most of our results do not depend on how the particles are out of equilibrium .
this broad insensitivity with respect to cosmological details is one of the key features to our work
. the order of our presentation will be as follows .
we begin with a discussion of the class of braneworlds and the energy scales and couplings that we are interested in .
we then present our constraint on this scenario for baryogenesis with dimension-6 operators in the low temperature regime .
finally , we summarize and conclude .
the simplest way to realize the proposal of geometrical proton decay suppression is for the `` brane '' on which we live to be a nontrivial scalar field profile in an extra dimension @xcite . on a smooth manifold ,
chiral fermions are localized in the presence of a scalar field profile with yukawa couplings to the fermion fields , resulting in a position - dependent effective mass term .
the simplest ansatz is to take a linearly varying scalar field , which results in chiral fermion zero modes with a gaussian wavefunction , localized about the point where the effective fermion mass vanishes @xcite . on an orbifold
@xcite one may impose special boundary conditions , which themselves ensure that the lowest mode is chiral , then with a scalar field which is forced to take up a kink configuration one obtains a fermion wavefunction which varies approximately as @xmath38 . by coupling the leptons and quarks differently to different scalar fields , one may separate the zero mode wave functions such that the wave function overlap factor @xmath39 in eq .
( [ eq : wavefunctionoverlap ] ) obtains an exponentially suppressed value .
as discussed in @xcite , it is not easy to induce _ large _ displacements of the points of localization away from special points such as the orbifold fixed points ( whose boundary conditions force the zero of the scalar field ) without introducing arbitrary scale parameters . in this paper
, we will not assume any particular mechanism to make these localization points ( which we will sometimes loosely call `` branes '' ) dynamical , but rather suppose that large dynamical displacements of the localization points ( branes ) from their final value are possible .
hence , we treat @xmath39 as merely a time - dependent function that is controlled by an exponential of the form @xmath40 where @xmath41 is a constant coefficient of order 1 and @xmath17 is a time - dependent function giving the distance between the quark and lepton localizations
. there may be additional contributions to @xmath39 varying as @xmath42 but these will be subleading unless @xmath43 .
we will see below that @xmath41 actually has to be greater than or equal to about @xmath44 in our scenario .
the baryogenesis scenario is then as follows : at some early time in the evolution of the universe , @xmath45 and the baryon number violating operator is suppressed only by the fundamental scale . by the time of nucleosynthesis , @xmath46 will have settled down to today s value of about 50 , consistent with proton decay constraints ( see appendix [ app : pprotection ] ) . for the purposes of our paper
, we will not need the specific details of how the quark - lepton separation is stabilized at its final value .
we will show that the short - distance physics prevents sufficient baryogenesis in the minimal model that we consider , except in the special case already mentioned .
since the zero mode quark and lepton wave functions are separated in the extra dimension and since the quarks and the leptons interact through the @xmath47 gauge group , these gauge fields must also propagate in the extra dimension , and the 4d gauge theory must come from dimensional reduction . similarly , since everything interacts with gravity , 4d gravity also arises from dimensional reduction .
we now derive bounds on the fundamental scale as well as the size of the extra dimensions by demanding perturbativity of gauge fields propagating in the extra dimensions and that einstein gravity be recovered . from gravitational dimensional reduction ,
given that there are a total of @xmath33 extra dimensions and one of them is asymmetric with length @xmath48 , the others having lengths @xmath49 , @xmath50 , we have @xmath51 where @xmath52 denotes the present time and @xmath53 is the reduced planck mass @xmath54 .
there is also a perturbativity constraint on the yang - mills coupling constant . in @xmath55 dimensions ( for example for a @xmath56-brane ) ,
the yang - mills coupling is @xmath57 where @xmath58 is a dimensionless number . by dimensional reduction to 4d
, we have the 4-dimensional dimensionless yang - mills coupling @xmath59 related to @xmath58 by @xmath60 if @xmath61 is to remain perturbative at time @xmath62 , we must have @xmath63 given that @xmath64 for the ew theory [ @xmath65 $ ] , we require @xmath66 if we have a scenario in which @xmath67 this is not an issue , but in general , this is constraining . now , let us consider absolute perturbativity of the five dimensional theory .
since the cutoff is set at @xmath7 , perturbativity requires @xmath68 using eq .
( [ eq:4dandhigher ] ) , we obtain @xmath69 for 5d perturbativity .
however , this requires the volume enhancement factor to be small . in the end , if we want to minimize this effect , we must have only a d4-brane ( _ i.e. _ @xmath70 ) .
note that our perturbativity constraint is for calculational feasibility rather than necessarily for fundamental physical reasons .
coming back to eq .
( [ gravityscale ] ) , if we impose that @xmath71 in which the gauge fields propagate is the largest extra dimension @xmath72 where the approximate equality is due to eq.([5dpertbound ] ) , one sees that @xmath73 to restate this , if the gauge fields come from dimensional reduction of the _ largest compact _ dimension , perturbativity requires @xmath74 hence , for this scenario to have @xmath75gev , there must be compact dimensions larger than the dimension in which the gauge fields propagate
. then we take , for simplicity , 5 large extra dimensions of length @xmath76 in which the gauge fields do not propagate , and one comparatively smaller extra dimension of length @xmath77 in which ( electroweak ) gauge fields do propagate . since in that case ,
@xmath78 we can choose @xmath79 furthermore , the experimental limit from neutron - antineutron oscillations forces @xmath80gev @xcite , therefore @xmath81 must be smaller than about @xmath82gev@xmath83 . in the spirit of the original motivation for large extra dimensions , namely in order to ameliorate the hierarchy problem , we take the smallest consistent value of @xmath7 and set it to @xmath84 gev . what is important to keep in mind from this section is eq .
( [ 5dpertbound ] ) .
since @xmath85 , the maximum value that @xmath86 can take is @xmath87 .
hence , proton decay constraint requires @xmath88 as claimed below eq .
( [ eq : defzeta ] ) .
furthermore , to `` unsuppress '' the dimension-6 operator , @xmath89 is required during baryogenesis .
this means that if the distance between the branes @xmath17 is equal to the time dependent size of the extra dimension @xmath90 , then @xmath91 during baryogenesis .
however , as shown in fig .
[ fig : branesetup ] , if the distance between the branes can be adjusted independently of the size of the extra dimension , this need not be true . finally , if we require that no k - k modes of the @xmath47 gauge group contribute to the baryogenesis dynamics , we must have @xmath92 .
this leads to the requirement that @xmath93 .
we will henceforth take @xmath94 to be a free ( time - dependent ) parameter , with a value @xmath95 with a setup in which proton decay and other @xmath5-violating processes are suppressed by the quark - lepton separation , the next question to ask is which processes dominated at the epoch when @xmath17 was smaller ( @xmath96 ) .
proton decay due to massive fermion exchange , which appears in the effective theory by operators suppressed as @xmath97 with @xmath98 , is the most `` difficult '' to suppress at the present epoch , since one requires @xmath99 at the present time up to factors of order 1 , and @xmath100 is bounded from above by @xmath101 if su@xmath102 is to remain perturbative in the bulk .
decay mediated by local interactions in the @xmath103 effective field theory , requiring the localized sm fermions to propagate through the bulk with a suppression of @xmath104 , might also be problematic if @xmath105 is small relative to @xmath106 , as it must be for a credible field theory model of localization . here
we would require @xmath107 today .
nonperturbative effects involving the extra dimensions should also be considered , for example virtual black holes or wormholes connecting the two branes .
however , given very reasonable assumptions for the topology of the corresponding instantons , their action is large enough for the effects to be negligible @xcite .
all other @xmath100-dependent effects turn out generically to be smaller , including the weak instantons which do not cause proton decay but might be important in baryogenesis ( appendix [ app : inst ] ) .
then , assuming that both suppression factors saturate the current bound , as @xmath100 shrinks from its present value the effective operators suppressed by @xmath104 grow much more quickly than those varying as @xmath108 . at the time of baryogenesis , for @xmath100
comparable to @xmath7 , the dominant processes ( all other things being equal ) will be those mediated by the sm fermions propagating through the bulk , for which the suppression is @xmath109 .
the goal now is to see if without an electroweak phase transition , we can create enough baryon asymmetry with the dimension-6 operator which is unsuppressed .
the defining characteristic of our class of baryogenesis scenarios is the leading baryon number - violating dimension-6 operator , one example of which is written in the gauge eigenstate basis as @xmath110\ ] ] where we have suppressed color indices , @xmath111 are family - dependent coefficients , @xmath112 are @xmath113 indices , summation of indices is implicit , and @xmath114 , @xmath115 and @xmath116 are constant real coefficients .
the wave function overlap integral that gives rise to @xmath39 will be taken to be @xmath117 during baryogenesis .
rotating the quarks and leptons into mass eigenstates , we have @xmath118 @xmath119 which corresponds to the usual flavor mixing matrices @xmath120 and @xmath121 .
this then gives @xmath122\ ] ] where @xmath123 note that due to the introduction of the dimension-6 operators , the global symmetry of the theory ( in the limit that the yukawa couplings vanish ) is different from the usual low energy analysis of the standard model .
hence , we have more than one rephasing invariant cp - violating complex parameter .
now consider for example the interference between the tree level and the one - loop diagrams shown in fig .
[ fig : interf ] , assuming for simplicity that @xmath124 and @xmath125 .
if the tree level amplitude is denoted as @xmath126 while the one loop amplitude is labeled as @xmath127 , we know that @xmath126 can be written as @xmath128 where @xmath129 is purely real , while @xmath127 will in general be of the form @xmath130 where @xmath131 and @xmath132 are real functions .
note that @xmath132 will in general arise if we can cut the propagators in the loop ( to make two tree level graphs ) and put them on shell simultaneously . for the @xmath0 conjugate reaction @xmath133 , but @xmath134 where in particular , the imaginary part arising from the cut propagators is not the complex conjugate of that for the @xmath0-conjugate reaction , since it derives from kinematics alone . in the boltzmann equations for the baryon number density
, the probability of interest will be @xmath135 \\ & = & 4 \langle r_1 ( i_2 + i_{\bar{2 } } ) \rangle \delta \end{aligned}\ ] ] where we have defined @xmath136\ ] ] and the @xmath137 represents spin averaging .
written out explicitly , this interference term is @xmath138 \ } + \mbox{parity odd}\ ] ] where the external momenta @xmath139 are labeled clockwise and constrained to satisfy @xmath140 , @xmath141 , and @xmath142 \left[\frac{(p_4 + k)^\beta } { ( p_4 + k)^2 -m_{d_{j}}^2 } \right ] \frac{1}{k^2-m_w^2 } \right\}\ ] ] note that @xmath143 vanishes only if the intermediate quark states can not go on shell .
if the intermediate quark states can go on shell , the expression becomes @xmath144 now , let us make some estimates .
consider the standard parameterization of @xmath145 .
suppose we choose the angles as @xmath146 , @xmath147 , @xmath148 , and a phase @xmath149 roughly consistent with experimentally known bounds .
we specify in part a yukawa texture by choosing the rotation matrix @xmath150\ ] ] and suppose @xmath151 . with a array of random ( real ) numbers of order 1 for the @xmath152 , we find the effective @xmath0 violating imaginary coefficient to be distributed as shown in fig .
[ fig : imaginarycoeffs ] violating coefficients from a random texture . ]
where we have summed over the indices @xmath153 as if rest of the transition probability did not depend on @xmath154 and @xmath155 .
as expected , the quantity @xmath156 ( analogous to the jarlskog invariant in the presence of the dimension-6 operators ) is around @xmath157 , much larger than the jarlskog invariant of @xmath158 .
we expect similar results if the @xmath152 are taken complex with random phases . as for the @xmath143 we have when the mandelstam invariant @xmath159 @xmath160 while for the top quark , @xmath161 note that @xmath162 can be time dependent since @xmath163 and @xmath94 can be time dependent ( anywhere between @xmath164 ) .
although @xmath165 may be time - dependent , we will take it to be approximately constant . and
@xmath166 gev for the standard thermal scenario .
] now , let us write down the boltzmann equation for our system of quarks and leptons . denoting the phase space density of particles in an frw cosmology with scale factor @xmath167 as @xmath168 , we have @xmath169 where @xmath170 is an operator generating an appropriate momentum integral ( with the measure @xmath171 $ ] ) , @xmath126 is the amplitude for the tree level process @xmath172 , and @xmath173 is associated with the one loop correction with intermediate states @xmath174 .
the amplitude @xmath175 is associated with the tree level amplitude of @xmath176 , where @xmath174 are the intermediate states in the calculation of @xmath127 , and the one loop correction to @xmath175 involving the intermediate states @xmath177 ( appearing in @xmath126 ) is denoted as @xmath178 . here , we have neglected the pauli blocking factors for the boltzmann collision integral .
although this is easily generalizable to include the pauli blocking factors and although it is not justified when calculating an accurate number density evolution , the neglect is justified when calculating an upper bound on the baryon numbers produced .
hence , for clarity of physics we shall drop them in this paper . because only the amplitudes with the fermion propagators cut in @xmath127 and @xmath179 contribute
, one can show that within the collision integral sum that @xmath180 .
note that the collision integral vanishes in equilibrium .
( a related work on this kind of boltzmann equation may be found in ref .
@xcite . )
thus one arrives at the baryon number density evolution equation @xmath181 \label{eq : genericbaryonevolve}\ ] ] where @xmath182 is the baryon number associated with reactant @xmath167 and the prime on the integral denotes that all the momenta are to be integrated over . note that for each baryon number associated with reactant @xmath167 to contribute to the baryon asymmetry , the intermediate and the final state must be out of equilibrium and not the initial reactant .
this is expected from unitarity and cpt which implies that @xcite @xmath183 where @xmath155 sums over all final states : when the probability is summed over all final states , there can not be any @xmath0-violating enhancement or subtraction of the probability that survives .
also , note that although navely one might think that the rhs of eq .
( [ eq : genericbaryonevolve ] ) would vanish , since the quantity inside the square bracket is odd under the exchange @xmath184 and the summation is over all @xmath185 , there is no cancellation because @xmath186 is also odd under the exchange @xmath184 .
finally , note that although we have only written down the scattering processes explicitly , the decay and inverse decay reactions have an analogous form .
now , instead of accounting for the possible contributions to the baryon asymmetry from all the reactants together , let us put bounds on the the maximum contribution that arises from each of the reactants .
as explained in appendix [ appendixboltzmann ] , we can derive from eq .
( [ eq : genericbaryonevolve ] ) @xmath187 where @xmath188 @xmath189 , @xmath190 and @xmath191 are temperature and mass scales , respectively , of the initial state ( or final state ) reactants , @xmath192 and @xmath193 are the masses of the particles in the loop , @xmath194 is the tree - level baryon number violating reaction rate which can be read off from table [ tab : bvrates ] . in the table
the function @xmath195 defined as @xmath196 is a time dilation function arising from the thermal averaging and @xmath197 is an function of order 1 also arising from thermal averaging ( @xmath198 is a modified bessel function of the second kind ) .
@xmath199 + @xmath200 & @xmath201 & @xmath202 & @xmath203 + @xmath204 & @xmath205 & @xmath206 & @xmath207 + @xmath208 & ' '' '' @xmath209 & @xmath210 & @xmath203 + @xmath211 & ' '' '' @xmath212 & @xmath213 & @xmath207 + using these and other results presented in appendix [ appendixboltzmann ] , we can place an upper bound on the baryon asymmetry generated by any single species out of equilibrium as @xmath214 _
{ \rm max}\ ] ] where the maximum is taken at any time during baryogenesis . note that this bound is independent of cosmology ( without specifying the time dependence of the scale factor @xmath215 ) as long as the usual boltzmann equations are valid .
let us first consider the quarks . instead of writing @xmath216 as the sum of rates for all the channels ,
it is more illuminating to consider separately the cases when @xmath217 is dominated by either the baryon number - conserving annihilation rate or the decay rate . the @xmath5-conserving annihilation rate is dominated by a t - channel quark annihilation into gluons , with the possible exception of the top quark for which the rate can be dominated by a weak scattering .
explicitly , we use @xmath218 @xmath219 is the appropriate quark mixing matrix element squared , and @xmath220 is a function encoding the effect of thermal averaging , similar to @xmath221 .
the @xmath5-conserving decay rate is dominated by a weak decay rate , which is enhanced for the top quark due to the longitudinal @xmath222 component going on shell .
we use @xmath223 @xmath224 explicitly , @xmath225 while @xmath226 , which implies that the weak decay rate is not much enhanced for the top quark through the goldstone emission .
given these rates , we list the leading bound @xmath227 ( _ i.e. _ the upper bound on the baryon fraction relevant to the case that the given reactions are the dominant contributions ) in table [ tab : quarkrates ] . the lepton sector is similar except with @xmath228 and no top quark enhancements .
in addition , the leading number density changing channel for the neutrinos may be coannihilations instead of self - annihilations .
@xmath229 + 1 & @xmath230 & ' '' '' @xmath231_{\rm max}(\times e_{t})$ ] + 2 & @xmath232 & ' '' '' @xmath233 _ { \rm max}(\times e_{t } ) $ ] + 3 & @xmath234 & ' '' '' @xmath235 + 4 & @xmath236 & ' '' '' @xmath237_{\rm max}(\times f_{t } ) $ ] + from the last section , we have @xmath238 tev which implies @xmath239 finally , we need to constrain the number density of the particles .
since the elastic scattering rates are at least as large as the total inelastic scattering rates , to absolutely forbid any phase transitions , we restrict the number density of particles to be smaller than the thermal density at @xmath240 gev , _
i.e. _ @xmath241 if relativistic or @xmath242 if nonrelativistic . now , consider the case when the branching ratio is approximated by ratio 3 in table [ tab : quarkrates ] .
for the top quarks out of equilibrium , we have @xmath243 while for all other reactants , we have @xmath244 where we have used the bottom quark mass of about 4.3 gev . for the top quark , because the upper bound here is smaller than what is needed for sufficient baryogenesis and since the same numerator appears in ratio 1 , the @xmath5-violating decay channel is ruled out for the top quark . for the bottom quark , since the leading decay channel has latexmath:[$ rules out sufficient baryogenesis .
all other channels also fail the tests under ratios 1 and 3 which means that the @xmath5-violating decay channel can not help in generating sufficient baryon asymmetry .
consider now ratios 2 and 4 .
ratio 4 requires for the top quark that @xmath246 if we assume that @xmath247 .
hence , ratios 3 and 4 completely rule out the top quark as a significant contributor to the baryon asymmetry .
let us now consider the scattering of the out - of - equilibrium particle @xmath248 with every other particle labeled as @xmath249 in view of ratio 2 .
it can easily be seen that ratio 2 is not very constraining , as it can in principle be made arbitrarily large .
hence , we can consider a more stringent version of the entropy bound corresponding to ratio 2 , eq .
( [ eq : rigoroustext ] ) which we rewrite as @xmath250 where @xmath251 is the entropy density , @xmath252 is the temperature of particle species @xmath154 , @xmath253 is the characteristic temperature of species @xmath154 inside the integral , @xmath254 is a kinematic function of order 1 , @xmath255 marks the beginning time of baryogenesis , and @xmath256 marks the end time of baryogenesis . instead of using the maximum contribution to the entropy , it is sufficient to consider any one contribution since we are placing an upper bound .
since we are guaranteed that the reaction changing the number density of @xmath248 is out of equilibrium , and there exists a contribution to the entropy from the weak interaction ( @xmath257 ) with the matching integral of the form @xmath258 ( since @xmath248 is necessarily out of equilibrium ) , we can write using eq .
( [ eq : entropyboltzmann ] ) @xmath259 where @xmath260 is another order 1 dimensionless function .
it is important to note that we have assumed the kinematic viability of the weak interaction in eq .
( [ eq : weakentropy ] ) . combining eq .
( [ eq : weakentropy ] ) and eq .
( [ eq : bbound ] ) , we have @xmath261 independently of @xmath94 . hence , with @xmath262 , sufficient baryon asymmetry generation seems feasible . a case in which eq .
( [ eq : nogo1 ] ) applies is when out of equilibrium @xmath2 quarks are interacting with the @xmath1 quarks at an effective temperature of 30 gev . in particular ,
if the baryon number conserving weak interaction channel is dominated by @xmath263 which gives the @xmath264 suppression , sufficient baryogenesis in this scenario seems possible .
we will return to this possibility later and constrain it further .
another possibility to evade eq .
( [ eq : nogo1 ] ) is to have the weak interaction be kinematically suppressed . in the case of out of equilibrium @xmath1 quark , the reaction @xmath265 has the kinematic suppression necessary to evade eq .
( [ eq : nogo1 ] ) in the limit of zero temperature since @xmath266gev while @xmath267gev . in table
[ tab : possible ] , we list all the weak interactions that are kinematically suppressed such that sufficient baryogenesis is possible with respect to the constraints thus far .
.[tab : possible ] the set of reactions evading the simplest bound .
these reactions are suppressed when the temperature falls below the threshold temperature .
[ cols="^,^,^,^,^",options="header " , ] according to table [ tab : possible ] , some weak interactions are kinematically suppressed only below the temperature of the qcd phase transition , which is assumed to occur at around @xmath268 gev .
these interactions hence really are not suppressed in the temperature range of interest and the baryogenesis bound is not evaded for these reactants .
hence , these are labeled irrelevant in table [ tab : possible ] .
only interactions involving @xmath269 then remain as viable candidates for the kinematically suppressed baryon number conserving channel , if we assume that its mixing with @xmath270 or @xmath271 is less than @xmath272 . if the neutrino sector is mixed , all the bounds return and there are no candidates that can generate sufficient baryon asymmetry .
, we are assuming that the neutrinos have mass . in that case
, there may generically be scalars that can produce dimension-5 operators .
however , if these scalars are sufficiently heavy , then their contribution will be negligible .
we do not analyze such cases , for which beyond the standard model particles actively participate in the short distance physics . ] in both cases where eq .
( [ eq : nogo1 ] ) allows sufficient baryogenesis , the baryon number - conserving weak interactions between the reactants contribute too little to the entropy to rule out baryogenesis . in such situations ,
let us consider whether the self - annihilation interaction contribution to the entropy is enough to dilute away the baryon asymmetry contribution . start with eq .
( [ eq : bbound ] ) and again approximate @xmath273 .
however , instead of eq .
( [ eq : weakentropy ] ) , consider@xmath274 which corresponds to the entropy bound arising from self - annihilation reaction contributions , where the contribution to the entropy from the species @xmath275 only begins once it goes out of equilibrium , at time @xmath276 , while the @xmath248 self - annihilation contribution persists throughout the entire duration @xmath277 $ ] of baryogenesis . here ,
the thermal averaged quantities ( such as @xmath278 ) have their temperature fixed at some value , characteristic of the integration interval .
the advantage of considering the bound arising from self - annihilation reactions is that they change the number density and are always possible at tree level if there is a neutrino lighter than ( or degenerate with ) @xmath279 .
we shall assume that the self - annihilation channel is open .
that is nonzero even in the zero neutrino mass limit @xcite . ]
suppose first that the @xmath280 term dominates over the second term .
then , using the elementary inequality @xmath281 we can write @xmath282\ ] ] where we have defined @xmath283 and @xmath284 which measures how much the @xmath249 particles can contribute to baryogenesis through interaction with out of equilibrium particle @xmath248 without contributing to the entropy .
this can be seen from noting that @xmath285 is the equilibrium distribution during @xmath286 $ ] by definition .
producing baryon asymmetry for a long time without producing any entropy ( i.e. letting @xmath287 approach @xmath256 ) is the best that can be done to circumvent the suppression of the dimension 6 operator . for sufficient baryogenesis ( @xmath288 )
, we thus require @xmath289 -1\ ] ] where@xmath290 if the @xmath291 dominates in eq .
( [ eq : selfann ] ) , one can similarly derive eq .
( [ eq : lemma1 ] ) , thereby showing that it is true in general .
now , noting that one can write @xmath292 and using eq .
( [ trivineq ] ) , one can write @xmath293 it follows that @xmath294 which is useful if @xmath295 and @xmath296 , since as we mentioned before , @xmath285 is the equilibrium distribution during the time interval @xmath297 $ ] .
hence , we shall use eq .
( [ eq : maineq ] ) with parameterized cosmology to put a bound on cosmology from the condition of sufficient baryogenesis .
we parameterize the unknown cosmology as @xmath298 @xmath299 and determine @xmath276 by the out of equilibrium condition @xmath300 \label{eq : outofeq}\ ] ] where @xmath301 is an @xmath302 constant and @xmath303 is the decay rate of @xmath249 . because the ratio of the decay rate @xmath303 to the expansion rate increases as a function of time , when the decay becomes the dominant chemical potential changing reaction , it can lead to an out - of - equilibrium number density of the @xmath249 _ only if _ the final states of the decay are not in equilibrium .
hence , the maximum above applies only if the decay products are out of equilibrium , and otherwise , only the self - annihilation should be taken into account .
furthermore , note that any choice of @xmath287 that we make in eq .
( [ eq : outofeq ] ) corresponds to a different choice of cosmology because @xmath287 sets the scale of expansion rate of the universe . the larger @xmath287 is , the slower is the expansion rate , and hence the slower the dilution rate of any accumulation of baryon asymmetry . as far as interpreting the parameter @xmath304 is concerned , it is shown in appendix [ appendixscaling ] that @xmath305 for most physical systems ( even with external sources such as the inflaton ) confined to 3 spatial dimensions , unless there exists a cooling mechanism different from the expansion of the 3 spatial dimensions .
however , with the presence of extra dimensions , the energy exchange with the bulk exotics may induce a more rapid cooling allowing for @xmath306 .
if @xmath307 determines the out of equilibrium time of @xmath249 , we find @xmath308 where @xmath309 , @xmath310 is the number of degrees of freedom contributing to the entropy at time @xmath287 , and the factor of @xmath311 accounts for ( the square of ) the degrees of freedom in the @xmath249 number density ( @xmath312 is a modified bessel function of the second kind ) .
we will assume that @xmath310 accounts for at least the 16 degrees of freedom of the gluons in the thermal bath .
as for @xmath311 , if @xmath249 for example is a quark , then it will be @xmath313 for the 3 color degrees of freedom ( the 2 helicity states were already accounted for , which means that if @xmath249 were weyl neutrinos , this should be divided by 4 ) .
if @xmath303 determines the out of equilibrium time of @xmath249 , then this quantity should be multiplied by @xmath314 for the relevant reactions in table [ tab : possible ] , a @xmath279 and a quark are always paired . for the quarks out of equilibrium , @xmath315 and @xmath316 : @xmath317 -1,\ ] ] @xmath318 these equations tell us that for all effective temperatures ( between @xmath319 and @xmath320gev ) and quark masses ( less than @xmath82gev ) relevant to our scenario , we have the conditions @xmath321 and @xmath322 , which allow nontrivial bounds to come from eq .
( [ eq : mainnumeq ] ) . from eq .
( [ eq : mainnumeq ] ) , it is easy to show that there are no value of @xmath323 and @xmath324 such that sufficient baryon asymmetry is generated .
hence , there are no scenarios in which there is sufficient baryogenesis through out of equilibrium quarks interacting with @xmath269 .
for the case of @xmath279 out of equilibrium and @xmath325 in equilibrium , @xmath326 and @xmath327 and we have a similar result : @xmath328 -1,\ ] ] @xmath329 again , it is easy to check that for all relevant effective temperatures ( between @xmath330 and @xmath320gev ) and quark masses ( less than @xmath331 gev ) , we have the conditions @xmath321 and @xmath332 which allow nontrivial bounds to come from eq .
( [ eq : mainnumeq ] ) . as before , eq .
( [ eq : mainnumeq ] ) rules out the possibility of sufficient baryogenesis for @xmath324 and @xmath323 .
hence , even neglecting the mixings , it seems nearly impossible to have sufficient baryogenesis with @xmath269 through the dimension-6 operator .
let us now return to the case where the baryon number conserving weak interaction channel is dominated by @xmath263 which was seen to have sufficient suppression for us to further consider the possibility of successful baryogenesis . in applying the reasoning that led to eq .
( [ eq : mainnumeq ] ) , we can improve the bounds compared to just eq .
( [ eq : mainnumeq ] ) by considering that the entropy also has a lower bound at temperature @xmath333 of @xmath334 where @xmath335 corresponds to the density of gluons and by putting @xmath100 inside the time integral and tying all the temperature scales together .
the bound then becomes ( similarly to before ) the smaller of either @xmath336 ) } \label{eq : alternate}\ ] ] or eq .
( [ eq : mainnumeq ] ) . the large enhancement for the @xmath2-@xmath1 interaction scenario in both eq .
( [ eq : mainnumeq ] ) and eq .
( [ eq : alternate ] ) comes from the long time @xmath287 during which baryogenesis occurs while @xmath249 is in equilibrium .
this leads to an accumulation of any net gain in baryon asymmetry over the entropy .
more explicitly , @xmath287 varies as @xmath337 ( or the equivalent factor from eq .
( [ eq : correction ] ) when appropriate ) which comes from the exponentially large ratio of the short distance physics scale to the expansion rate scale when @xmath249 goes out of equilibrium
. the expansion rate must be exponentially small when @xmath249 goes out of equilibrium ( if at all ) because the self - annihilation reaction must maintain equilibrium even in the nonrelativistic regime in which the number densities are exponentially suppressed .
the resulting bound with @xmath338gev and @xmath339gev is shown in fig .
[ fig : almostenough ] .
as one can see , even with such an enhancement , there is insufficient baryogenesis ( or marginally enough ) in most cases . even for the marginal cases
, we do not expect the scenario to be realizable given the conservative nature of the bound .
other scenarios with small @xmath340 weak interactions while the heavier particle is out of equilibrium ( _ i.e. _ out of equilibrium @xmath2 interacting with @xmath341 , as well as out of equilibrium @xmath1 interacting with @xmath342 ) fail to generate sufficient baryogenesis for similar reasons .
a conservative upper bound for the baryon asymmetry generated by out of equilibrium @xmath2 quarks interacting with @xmath1 quarks . the parameters @xmath343 and @xmath304 parameterize various possible nonstandard cosmologies .
a value @xmath344 corresponds to inflating cosmologies and in the absence of abnormal cooling @xmath345 .
this scenario does not generate enough baryon asymmetry . ] if we do not assume thermal equilibrium initial conditions ( before relevant species go out of equilibrium ) , then as before we may consider the @xmath2 being in equilibrium while the @xmath1 quark is out of equilibrium . according to our explanation of the enhancement for this scenario before , the larger the mass of the particle that stays in equilibrium until temperature drops to around @xmath346 gev , the smaller the exponentially suppressed density of nonrelativistic particles , and then as long as the self - annihilation determines the out of equilibrium time @xmath287 ,
the longer is the time @xmath287 over which baryon number is generated , leading to a larger _ accumulation _ of baryon asymmetry . since @xmath347 in this scenario , we thus expect larger accumulation of baryon asymmetry over the entropy in this scenario .
in fact , the self - annihilation reaction becomes small by the temperature @xmath348 gev to the extent that it no longer controls the out of equilibrium time ( instead it is determined by the @xmath2-decay ) .
the competing entropy generation is suppressed because of @xmath349 for the weak interactions , and other reactions generating entropy are insufficient as well in diluting the baryon asymmetry . for completeness
, one should note that since @xmath350 in this scenario there is a contribution to the entropy not accounted for by eq .
( [ eq : mainnumeq ] ) and eq .
( [ eq : alternate ] ) .
this comes from the fact that @xmath2 undergoes out - of - equilibrium decay to @xmath1 even though @xmath2 has a chemical equilibrium distribution .
that is because @xmath1 by definition is out of equilibrium and thus can not give the inverse decay reaction necessary for the @xmath2 decay to conserve entropy .
however , we can check explicitly this contribution to the entropy contribution is negligible , even though it is proportional to the integration time @xmath287 , because of the smallness of the decay rate .
also , since the @xmath1 is out of equilibrium , the reaction of @xmath2-@xmath351 annihilation into @xmath1-@xmath352 will also contribute to the creation of new entropy .
however , one can check that his contribution is also insufficient to dilute away the baryon asymmetry generated .
the numerical bounds shown for this scenario in fig .
[ fig : enough ] indicate that this scenario is viable at the level of our analysis .
a conservative upper bound for the baryon asymmetry generated by out of equilibrium @xmath1 quarks interacting with @xmath2 quarks .
the parameters @xmath343 and @xmath304 are as before . for nonstandard cosmologies ,
sufficient baryogenesis may be possible .
however , as explained in the text , the out - of - equilibrium condition seems difficult to maintain . ]
the most obvious difficulty with this scenario is that another mechanism is required to set the initial density of @xmath2 quarks to be much larger than that of @xmath1 , as well as to maintain that hierarchy .
for example , for temperatures far above @xmath353 , the gluons which keep @xmath2 quarks in equilibrium will tend to produce a large number of @xmath1 quarks that will be close to chemical equilibrium .
since a high initial temperature is crucial , some mechanism besides just the expansion of the universe is required to keep @xmath1 out of equilibrium while is @xmath2 in equilibrium . even if the @xmath1 densities could somehow be kept small , since the self - annihilation cross sections for @xmath2 and @xmath1 are approximately the same , if the cosmological expansion is responsible for the out - of - equilibrium condition , then it must delicately tune itself to be smaller than the @xmath2 scattering rate while larger than that of the @xmath1 s . however , our `` model - independent '' analysis does not rule out this scenario , and it may be viable in a complicated and fortuitous setting . as for the scenario in which @xmath2 quarks
are in equilibrium while @xmath341 quarks are out of equilibrium , it will be even more difficult ( if not impossible ) to keep @xmath341 quarks out of equilibrium for the required `` exponentially long '' duration @xmath287 , because the gluons which are supposed to be in equilibrium have one of their strongest equilibration reactions through @xmath341-@xmath354 creation . can not have approximately equilibrium distribution if the particles with which @xmath355 interacts _ most strongly _ are not in equilibrium . ] for the scenario with other quarks ( besides the top and the bottom ) in equilibrium during baryogenesis , the exponential enhancement of the duration of baryogenesis is too weak to allow sufficient baryogenesis .
hence , remarkably , only one scenario , namely out of equilibrium @xmath1 quarks interacting with equilibrium @xmath2 quarks , may generate sufficient baryon asymmetry with the dimension 6 operator if the out of equilibrium mechanism can be appropriately engineered .
however , it would be surprising if such a realistic scenario can be constructed .
we have studied a new scenario of baryogenesis involving dimension-6 baryon number - violating operators that would generically arise from integrating out ultraviolet degrees of freedom in the context of the standard model .
the question to be answered in this scenario was whether a time - dependent dimensionless suppression coefficient for the dimension-6 operator can allow sufficient baryogenesis between the effective temperatures of @xmath356gev and @xmath320gev , if the only fields that are allowed to participate in the short distance physics are those of the standard model .
the upper bound on the temperature is motivated from the cosmological bound on kk graviton decay while the lower bound is purely for our calculational feasibility ( related to the qcd phase transition ) .
the time - dependent suppression coefficient was motivated from the generic possibility that with the arkani - hamed - schmaltz mechanism of suppressing proton decay , the proton decay suppression factor can be time - dependent .
unfortunately , if the fundamental scale setting the size of the unsuppressed dimension-6 operator is forced up to values @xmath357tev by the bound from neutron - antineutron oscillations , the branching ratio of @xmath5-violating interactions to @xmath5-conserving interactions is set by the quantity @xmath358 .
we have shown this to be too small to allow sufficient baryogenesis for most cases .
another way to view this bound is that as the dimension-6 operator creates baryon asymmetry , the same reactants participate in creating entropy through @xmath5-conserving operators .
the competition between the two types of reactions is almost always dominated by the @xmath5-conserving one .
hence , within the perturbative setting where the 4d effective action is valid , sufficient baryogenesis in this scenario is in most cases impossible almost independently of cosmology .
there is one scenario which can not simply be ruled out by the general analysis .
this involves out of equilibrium @xmath1 quarks interacting with equilibrium @xmath2 quarks from a temperature of about @xmath356gev until a temperature of @xmath320gev .
the reason why this scenario evades the suppression is that the primary entropy generation channel is through the @xmath2-@xmath351 self - annihilation into @xmath1-@xmath352 instead of the @xmath2-@xmath1 weak interaction , since @xmath359 .
furthermore , given that nonstandard cosmology of braneworlds may give a slow expansion rate , one can have baryogenesis persist for a long time period , giving a large integrated pileup .
however , given that @xmath1 quarks are produced efficiently both by the gluons that keep the @xmath2 quarks in equilibrium as well as the @xmath2 quarks which undergo decay , setting up the necessary hierarchy in densities to keep @xmath1 out of equilibrium while @xmath2 in equilibrium seems difficult .
however , this scenario does pass the general tests not relying upon the specifics of the out of equilibrium mechanism .
since neutrinos are almost massless and have vanishingly small tree level annihilation cross section in the limit of small @xmath360 , one would navely think that this channel might be weak enough for the baryon number - violating operator to compete with it .
unfortunately , this turns out not to be the case .
it is interesting to note that even with just the standard model fields , the presence of nonrenormalizable operators implies the existence of additional @xmath0-violating rephasing invariants beyond the jarlskog parameter . in any theory ,
the number of physical @xmath0-violating phases is given by the number of couplings in the lagrangian that are allowed to be complex , minus the dimension of the symmetry group which describes phase redefinitions of the fields : introducing new effective operators into the sm automatically creates more @xmath0-violating invariants , even if all the couplings of the new operators are real in some basis .
hence , it is not the smallness of the jarlskog invariant that prevents the success of our baryogenesis scenario .
it is merely the fact that unless neutron - antineutron oscillations are suppressed by some means other than the fundamental scale , the phenomenologically acceptable scale of @xmath361tev is still too large for dimension-6 operators to play a significant role for baryogenesis below the temperature of the electroweak phase transition . as with any `` no go '' arguments
, there are many loopholes in our conclusion ( besides the one remotely possible scenario that we already mentioned ) .
first , we neglected any possible nonperturbative physics .
the 4d coupling constants are inversely proportional to the volume of the extra dimensions in which the gauge fields propagate . thus ,
if the volume of the extra dimensions was small at some epoch in the early universe , then nonperturbative physics may dominate ( see appendix [ app : inst ] ) , given that the perturbative contributions to @xmath362 are insufficient in our scenario .
for example , the @xmath363 instanton effects may become unsuppressed , giving rise to a scenario similar to that of sphaleron transitions at the electroweak phase transition .
also , for effective temperatures below about @xmath346 gev , non - perturbative qcd effects will be relevant , since the quark degrees of freedom become confined .
although this is not likely to change the suppression of the dimension-6 operator , the kinematics and matrix elements will be very different from our perturbative calculation done in the regime of deconfined quarks .
secondly , we have neglected all effects of the kk modes ; or , in the higher dimensional picture , we have neglected the fact that the space may be inhomogeneous in the higher dimensions .
for example , one may envisage a scenario in which there is a phase transition which localizes the wave functions of the quarks and the leptons
. a proper description of quark - lepton separation in higher dimension will require analysis beyond the zero mode .
however , this type of effect will most likely play a role in enhancing the nonequilibrium condition rather than changing the baryon number - violating branching ratio .
hence , our conclusion is most likely robust with respect to this assumption .
thirdly , we have neglected all cpt violation effects that must exist because of the time dependence of the quark - lepton separation .
if the time scale associated with the quark - lepton separation process is very short , then there may be significant contributions from these effects to enhance the baryon number violating channels .
however , such models will probably be severely constrained by the restriction on bulk graviton production , just as the reheating temperature is severely constrained .
hence , although there are loopholes and caveats to our `` no go '' claims , it seems fair to conclude that with only perturbative physics , degrees of freedom beyond the standard model fields must play a significant role in baryogenesis if the effective temperature of the universe never exceeds 30 gev ( which is a very conservative upper bound for the maximum temperature in models with large extra dimensions , if the fundamental scale is to be accessible to collider experiments ) .
the requirement of beyond the standard model field content is not an obvious statement , considering that we hardly impose any restrictions on the cosmology and that there is an unsuppressed dimension-6 operator as well as an effectively large @xmath0 violating phase available for baryon asymmetry generation .
of course , our `` no go '' claims would be significantly relaxed if the dimension-9 operator responsible for neutron - antineutron oscillations is suppressed by some symmetry mechanism which is introduced by hand to supplement the geometrical suppression of proton decay operators . in closing , note that even if the fundamental scale can be lowered by evading the bound from neutron - antineutron oscillation , there is a significant challenge in this type of scenario as the quarks ( and/or leptons ) must be forced out of equilibrium at rather high temperatures . to accomplish this ,
it is possible to lower the planck scale during baryogenesis by having a small extra dimensional volume during that period .
however , there are severe restrictions coming from bulk graviton production in such cases .
furthermore , it is an extremely difficult challenge to find a time dependent potential for the scalar fields localizing the quarks and the leptons , such that the initial quark - lepton separation departs sufficiently from today s equilibrium value in a natural manner .
we leave variations on our model that may generate sufficient baryon asymmetry to future studies .
we thank aaron grant , nima arkani - hamed , misha shaposhnikov , lisa everett , rocky kolb , antonio riotto , marty einhorn , gordy kane , antonio masiero , tony gherghetta , alexander dolgov , karim benakli , gianfrancesco giudice and rula tabbash for helpful conversations and boris k " ors for correspondence .
this research was supported in part by doe grant de - fg02 - 95er40899 , task g.
to find the minimum value of @xmath17 consistent with proton stability , we consider the decay process in the four - dimensional low energy theory , where we have argued that four - fermion operators give the largest contribution .
( processes that can not be written as four - fermion effective operators , for example wormholes , are negligible for the value of @xmath17 that we find is needed . )
the proton decay rate calculation for these operators is analogous to muon decay in the fermi theory , giving an estimate of@xmath364 for the `` partial lifetime '' of decay into any allowed channel , where @xmath8 is a dimensionless constant of order 1 .
mode - dependent experimental lower limits on the proton lifetime @xcite require @xmath365 years for some modes , which would translate into in the absence of the suppression factor @xmath39 .
thus @xmath366 where the range quoted is for a fundamental scale ranging from 100 to @xmath367tev .
the conclusion is that about 50 e - foldings of suppression are required at the present epoch .
this is easy to obtain if @xmath368 and @xmath369 is not too small relative to @xmath7 , but is marginally consistent with perturbativity of su@xmath370 , which begins to break down at @xmath371 ( recall that @xmath372 ) , for proton decay mediated by states of mass @xmath7 , for which @xmath373 with @xmath374 of order 1 .
one should also consider nonperturbative processes which may affect @xmath5 violation and baryogenesis @xcite . even without the introduction of @xmath5-violating operators ,
the baryon number current is only conserved up to a su@xmath375 anomaly , and it is well - known that topologically nontrivial gauge configurations can lead to a change in @xmath376 , either by quantum tunneling via instantons , or thermal excitation via sphalerons .
anomaly is also nonvanishing , but there are no nontrivial u@xmath4 gauge configurations analogous to the instanton or sphaleron . ] the rates for @xmath5 violation by instantons and sphalerons also have exponential suppression factors @xmath377 and @xmath378 respectively , where @xmath379 is the sphaleron energy given by @xmath380 . if the total length of the fifth dimension @xmath90 were to change with time then , assuming a constant five - dimensional gauge coupling @xmath381 , these processes could become important . given the relationship @xmath382 for the weak gauge coupling , instantons would be exponentially larger for smaller @xmath90 ; the dependence is just @xmath383 , so the instanton enhancement at small @xmath90 would shadow the increase in processes mediated by massive states .
however , it can easily be checked that the instanton rate today is many orders of magnitude smaller than the bound on such perturbative processes , therefore instantons can never have dominated over the perturbative operators , unless the perturbative operators were further suppressed by many orders of magnitude ( for example , if they vanished by an exact discrete symmetry ) .
the sphaleron rate is more difficult to estimate , since in addition to the temperature dependence one would have to model how the higgs v.e.v .
changes in response to changing @xmath90 . on the simplest assumption of a constant v.e.v .
the sphaleron rate is actually independent of @xmath90 and , as in four dimensions , depends just on @xmath384 .
this is a counterexample to the claim that all processes are exponentially suppressed with increasing @xmath100 ; it is understood by noting that the formation of sphaleron configurations is a classical , thermally - activated process , in contrast to quantum effects which are suppressed by a massive propagator or by the euclidean action .
unsuppressed sphaleron transitions `` wash out '' any baryon asymmetry generated at temperatures above the electroweak scale ; number is generated at high scales .
] on the assumption of a very low maximum temperature of the universe , which is required in low - scale models , sphaleron effects will be completely negligible .
consider first the entropy of one species of particles . divide the phase space of particles into cells and
let the cells be numbered @xmath385 where each cell contains @xmath386 identical particles .
let @xmath387 be the phase space volume that can be occupied by one particle in the cell @xmath388 .
the total phase space of the @xmath389th cell is @xmath390 if one assumes @xmath391 and negligibly correlated distributions .
the total phase space volume is @xmath392 and the definition of entropy is @xmath393 . using the approximation @xmath394
, we can express the entropy as @xmath395\\ & \approx & \int \frac{d^{3}pd^{3}x}{(2\pi ) ^{3}}f(p , x)\ln \left[\frac{e}{f(p , x)}\right]\end{aligned}\ ] ] where @xmath396 gives the number density of particles .
hence , one can define the entropy current @xmath397\frac{p^{\mu } } { p^{0}}\ ] ] where we define the entropy density as usual as @xmath398 now , using the metric @xmath399 one has @xmath400 and hence the entropy density boltzmann equation becomes @xmath401\ln [ f_{f}]\ ] ] where @xmath402 corresponds to @xmath388th collision operator for the particle species @xmath403 in the boltzmann equation and @xmath404 represents the appropriate momentum phase space integral . neglecting @xmath29 violations
which are not important for the total entropy generation , we can write the collision terms in terms of s - matrix amplitudes as @xmath405\ln [ f_{f } ] & = & \sum _ { i}\int d\,\gamma ( 2\pi ) ^{4}\delta ^{(4)}(\textrm{momenta})\times \nonumber \\ & & ( f_{a_{i}}f_{b_{i}}-f_{c_{i}}f_{d_{i}})|m(c_{i},d_{i};a_{i},b_{i})|^{2}\ln [ \frac{f_{a_{i}}f_{b_{i}}}{f_{c_{i}}f_{d_{i}}}]+\nonumber \\ & & \sum _ { i}\int d\,\gamma ( 2\pi ) ^{4}\delta ^{(4)}(\textrm{momenta})\times \nonumber \\ & & ( f_{e_{i}}-f_{g_{i}}f_{h_{i}})|m(g_{i},h_{i};e_{i})|^{2}\ln [ \frac{f_{e_{i}}}{f_{g_{i}}f_{h_{i}}}]+\nonumber \\ & & \sum _ { i}\int d\,\gamma ( 2\pi )
^{4}\delta ^{(4)}(\textrm{momenta})\times \nonumber \\ & & ( \textrm{initial states}-\textrm{final states})|m|^{2}\ln [ \frac{\textrm{initial states}}{\textrm{final states } } ] + ... \nonumber \\ & & + \delta c\label{eq : entropyboltzmann } \end{aligned}\ ] ] where the phase space integral @xmath406 is over all particles and momenta , and @xmath407 corresponds to the entropy change produced by the decay of a classical field configuration . if a zero - temperature classical field configuration ( such as an inflaton vev ) decays to particles @xmath408 and @xmath409 and the bogoliubov coefficient for the particle production is @xmath410 ( where the physical momentum is @xmath411 ) , then we have the effective collision operator @xmath412\ln [ f_{x_{1}}]+\int d\,\gamma c_{i}^{x_{2}}[f_{x_{2}}]\ln [ f_{x_{2}}]=-\int \frac{d^{3}p}{(2\pi ) ^{3}}\ln [ f_{x_{1}}f_{x_{2}}]\frac{d}{dt}|\beta _ { pa}|^{2}\ ] ] contributing to the entropy density boltzmann equation . with our expression for the entropy , one can prove the following little theorem : if baryons are generated through particle @xmath403 which is out of equilibrium undergoing arbitrary reactions , then @xmath413 and @xmath414 _ { \rm max}\ ] ] where @xmath415 @xmath416 is the effective energy scale which characterizes the kinetic energy of the on shell particles in the loop , is approximately @xmath417 where @xmath190 and @xmath191 are temperature and mass scales of the initial reactants and @xmath192 and @xmath193 are the masses of the particles in the loop . ]
@xmath418 is the loop integral suppression factor , @xmath165 is the higgs vev of about @xmath166 gev , and @xmath194 is the tree - level baryon number violating reaction rate .
the maximum in eq .
( [ eq : roughbound ] ) is taken during when most of the baryogenesis is taking place , and @xmath419 corresponds to the reaction rate of @xmath403 which does not violate baryon number . the argument validating this claim goes as follows .
starting from eq .
( [ eq : genericbaryonevolve ] ) , one can use the momentum conserving @xmath156-function and the form of the equilibrium distribution to obtain eq .
( [ eq : rigorous ] ) if one assumes that the baryon number - violating operator coupling is of order 1 ( maximal ) .
now , consider the @xmath403 evolution . defining @xmath420 , we can write the @xmath5-conserving part of the boltzmann equation as @xmath421 where @xmath422 are baryon number - conserving reaction rates , @xmath423 are equilibrium number densities per comoving volume , @xmath424 are the ratios of final state number densities to initial state equilibrium number densities that usually appear in the boltzmann equations , and @xmath425 is the external source for @xmath403 .
since we are assuming maximal out of equilibrium , the equilibrium terms involving reverse reactions can be neglected .
hence , we can write @xmath426 which governs @xmath427 neglecting the baryon number violating terms . hence ,
we can write @xmath428 furthermore , we can write @xmath429_{\rm max } \left(\int _ { t_{i}}^{t}dt'\,\delta c(t')a^{3}(t')-n_{f}(t)+n_{f}(t_{i})\right)\ ] ] from eq . (
[ eq : entropyboltzmann1 ] ) , we see that @xmath430 hence , we find eq .
( [ eq : roughbound ] ) , as promised .
here we discuss the physical range of @xmath431 in @xmath432 .
the generalized first law of thermodynamics tells us @xmath433 where @xmath434 is the external source / sink of energy for the photons and for other relativistic fluids in thermal equilibrium with the photons .
here we have separated the 3 dimensional pressure @xmath435 from the component associated with the extra dimension , whose contribution to @xmath436 we write as @xmath437 . if we assume a constant `` effective '' average equation of state @xmath438 , set @xmath439 , and assume that @xmath440 , then we find , trivially
, @xmath441 for photons and other relativistic particles in thermal equilibrium confined to 3 spatial dimensions , the equation of state is @xmath442 and @xmath443 .
this implies @xmath444 which implies that @xmath445 unless the source term @xmath446 is significant . if @xmath447 , then @xmath448 , otherwise @xmath306 .
an example of a situation in which @xmath448 is when the bulk fields are in thermal equilibrium with the brane fields through @xmath449 where @xmath450 is the bulk pressure and @xmath451 is the scale factor for the extra dimension .
lykken , phys .
d * 54 * ( 1996 ) 3693 [ hep - th/9603133 ] .
n. arkani - hamed , s. dimopoulos and g r. dvali , phys .
b * 429 * ( 1998 ) 263 [ hep - ph/9803315 ] , phys .
d * 59 * ( 1999 ) 086004 [ hep - ph/9807344 ] .
i. antoniadis , n. arkani - hamed , s. dimopoulos and g.r .
dvali , phys .
b * 436 * ( 1998 ) 257 [ hep - ph/9804398 ] .
t. banks and l.j .
dixon , nucl .
b * 307 * ( 1988 ) 93 ; g. gilbert , nucl .
b * 328 * ( 1989 ) 159 .
d. e. groom _ et al . _
[ particle data group collaboration ] , eur .
j. c * 15 * ( 2000 ) 1 .
k. benakli and s. davidson , `` baryogenesis in models with a low quantum gravity scale , '' phys .
d * 60 * , 025004 ( 1999 ) [ hep - ph/9810280 ] .
g. aldazabal , s. franco , l. e. ib ' a~ nez , r. rabadan and a.m. uranga , jhep * 0102 * ( 2001 ) 047 [ hep - ph/0011132 ] ; l.e .
ib ' a~ nez , f. marchesano and r. rabadan , `` getting just the standard model at intersecting branes , '' arxiv : hep - th/0105155 . l.j . hall and d.r .
smith , phys .
d * 60 * ( 1999 ) 085008 [ hep - ph/9904267 ] ; m. fairbairn , phys .
b * 508 * ( 2001 ) 335 [ hep - ph/0101131 ] ; s. hannestad , phys .
d * 64 * ( 2001 ) 023515 [ hep - ph/0102290 ] .
m. fairbairn and l. m. griffiths , `` large extra dimensions , the galaxy power spectrum and the end of inflation , '' arxiv : hep - ph/0111435 . s. cullen and m. perelstein , phys .
* 83 * ( 1999 ) 268 [ hep - ph/9903422 ] ; v. barger , t. han , c. kao and r. j. zhang , phys .
* b461 * , 34 ( 1999 ) [ hep - ph/9905474 ] ; s. hannestad and g. raffelt , phys .
* 87 * ( 2001 ) 051301 [ hep - ph/0103201 ] , `` stringent neutron - star limits on large extra dimensions , '' arxiv : hep - ph/0110067
giudice , r. rattazzi and j.d .
wells , nucl .
* b544 * , 3 ( 1999 ) [ hep - ph/9811291 ] ; e.a .
mirabelli , m. perelstein and m.e .
peskin , phys .
* 82 * ( 1999 ) 2236 [ hep - ph/9811337 ] ; t. han , j.d . lykken and r.j .
zhang , phys .
d * 59 * ( 1999 ) 105006 [ hep - ph/9811350 ] ; j.l .
hewett , phys .
* 82 * ( 1999 ) 4765 [ hep - ph/9811356 ] ; i. antoniadis , phys .
b * 246 * ( 1990 ) 377 ; i. antoniadis , k. benakli and m. quiros , phys .
b * 331 * ( 1994 ) 313 [ hep - ph/9403290 ] , phys .
b * 460 * ( 1999 ) 176 [ hep - ph/9905311 ] ; p.
nath and m. yamaguchi , phys .
d * 60 * ( 1999 ) 116004 [ hep - ph/9902323 ] ; m. masip and a. pomarol , phys .
d * 60 * ( 1999 ) 096005 [ hep - ph/9902467 ] ; p.
nath , y. yamada and m. yamaguchi , phys .
b * 466 * ( 1999 ) 100 [ hep - ph/9905415 ] ; t.g .
rizzo and j.d .
wells , phys .
d * 61 * ( 2000 ) 016007 [ hep - ph/9906234 ] ; c.d .
carone , phys .
d * 61 * ( 2000 ) 015008 [ hep - ph/9907362 ] .
l.m . krauss and f. wilczek ,
* 62 * ( 1989 ) 1221 .
r. blumenhagen , b. k " ors , d. l " ust and t. ott , nucl . phys .
b * 616 * ( 2001 ) 3 [ hep - th/0107138 ] ; r. blumenhagen , b. k " ors and d. l " ust , jhep * 0102 * ( 2001 ) 030 [ hep - th/0012156 ] ; r. blumenhagen , l. g " orlich , b. k " ors and d. l " ust , jhep * 0010 * ( 2000 ) 006 [ hep - th/0007024 ] . s. davidson , m. losada and a. riotto , phys . rev
. lett .
* 84 * ( 2000 ) 4284 [ hep - ph/0001301 ] .
dvali and g. gabadadze , phys .
b * 460 * ( 1999 ) 47 [ hep - ph/9904221 ] .
dienes , e. dudas and t. gherghetta , nucl .
b * 537 * ( 1999 ) 47 [ hep - ph/9806292 ] .
adams , g.l .
kane , m. mbonye and m.j .
perry , int .
j. mod .
phys . a * 16 * ( 2001 ) 2399 [ hep - ph/0009154 ] .
a. masiero , m. peloso , l. sorbo and r. tabbash , phys .
d * 62 * ( 2000 ) 063515 [ hep - ph/0003312 ] .
r. allahverdi , k. enqvist , a. mazumdar and a. perez - lorenzana , nucl .
b * 618 * ( 2001 ) 277 [ hep - ph/0108225 ] ; a. mazumdar and a. perez - lorenzana , `` affleck - dine baryogenesis in large extra dimensions , '' arxiv : hep - ph/0103215 .
a.d . sakharov , pisma zh .
teor . fiz .
* 5 * ( 1967 ) 32 [ jetp lett .
* 5 * ( 1967 ) 24 ] .
k. akama , `` an early proposal of brane world , '' lect .
notes phys .
* 176 * ( 1982 ) 267 [ hep - th/0001113 ] ; v.a .
rubakov and m.e .
shaposhnikov , phys .
b * 125 * ( 1983 ) 136 .
h. georgi , a.k .
grant and g. hailu , `` chiral fermions , orbifolds , scalars and fat branes , '' phys .
d * 63 * , 064027 ( 2001 ) [ hep - ph/0007350 ] .
dolgov , `` baryon asymmetry of the universe and violation of the thermodynamic equilibrium , '' pisma zh .
* 29 * , 254 ( 1979 ) .
kolb and s. wolfram , nucl .
b * 172 * , 224 ( 1980 ) [ erratum - ibid .
b * 195 * , 542 ( 1980 ) ] .
crewther , j. finjord and p. minkowski , `` the annihilation process neutrino anti - neutrino @xmath452 gamma gamma with massive neutrinos in cosmology , '' nucl .
b * 207 * , 269 ( 1982 ) .
kuzmin , v.a .
rubakov and m.e .
shaposhnikov , phys .
b * 155 * ( 1985 ) 36 .
rubakov and m.e .
shaposhnikov , usp .
nauk * 166 * ( 1996 ) 493 [ phys .
* 39 * ( 1996 ) 461 ] [ hep - ph/9603208 ] .
c.t . hill and p. ramond , nucl .
b * 596 * ( 2001 ) 243 [ hep - th/0007221 ] . | we study a new baryogenesis scenario in a class of braneworld models with low fundamental scale , which typically have difficulty with baryogenesis .
the scenario is characterized by its minimal nature : the field content is that of the standard model and all interactions consistent with the gauge symmetry are admitted .
baryon number is violated via a dimension-6 proton decay operator , suppressed today by the mechanism of quark - lepton separation in extra dimensions ; we assume that this operator was unsuppressed in the early universe due to a time - dependent quark - lepton separation .
the source of @xmath0 violation is the ckm matrix , in combination with the dimension-6 operators .
we find that almost independently of cosmology , sufficient baryogenesis is nearly impossible in such a scenario if the fundamental scale is above 100 tev , as required by an unsuppressed neutron - antineutron oscillation operator .
the only exception producing sufficient baryon asymmetry is a scenario involving out - of - equilibrium @xmath1 quarks interacting with equilibrium @xmath2 quarks . |
we review the recent work on the intercalation chemistry of iron chalcogenide superconductors . given that they exhibit zero resistance for electrical currents below a critical temperature ( @xmath1 ) ,
superconductors hold great promise in our future energy needs.@xcite furthermore , superconductors have been proposed for devices that stabilize the electrical power grid by storing energy mechanically in flywheels or electromagnetically in toroidal magnets.@xcite much as previously known superconducting materials , the iron - based compounds have a role to play in our future energy needs.@xcite the iron chalcogenide superconductors remain topical due to their versatile solid state chemistry that allows new materials to be discovered and their ability to be isolated as single layers .
furthermore , the highest @xmath1 so far observed in any iron - based system has been in single layered fese where reports vary from 65 k to 100 k.@xcite just like the high-@xmath1 cuprates , the iron - based systems appear to be unconventional in their superconducting mechanism@xcite and this raises more hope that solid state chemists will continue to make significant discoveries in this field .
although similar in many respects to the iron - based pnictides , the chalcogenides do show considerable differences in their physical and chemical properties . while the arsenide phases are held together by ionic forces between the cationic layers , @xmath2 ( lao)@xmath3 and ba@xmath4 , and the anionic ( feas)@xmath5 layers , the chalcogenides can be held by van der waals interactions alone .
these comparatively weaker interactions make the fe@xmath6 layers , where @xmath7s@xmath8 and se@xmath8 , ideal hosts for intercalation chemistry such as alkali metal insertion along with various other types of guest species such as ammonia .
since the pnictides do not express intercalation chemistry , we have left them out altogether in this review .
the reader curious about the major differences between chalcogenides and pnictide superconductors can find several reviews written on these two major categories.@xcite we review the chemical techniques for inserting ionic and molecular species into fe@xmath6 hosts and their resulting physical properties .
an example of the different layered structures we will be reviewing and their intercalation chemistry is presented in the schematic of figure [ fig_intro ] .
we will examine three major methods for intercalating guest species , or _ intercalates _
, which include insertion of 1 . )
electropositive metals in liquid ammonia , 2 . ) metals and organic amines , and 3 . )
extended hydroxides through hydrothermal conditions .
each method plus post - synthetic treatment can lead to guest - host compounds with different structures as shown in figure [ fig_intro ] . before presenting the latest studies on the intercalation of iron - based materials , we briefly review the history of intercalation chemistry of superconductors such as c@xmath9 and metal dichalcogenides .
it is also instructive to present the crystal chemistry and physical properties of the simple binary iron chalcogenides fes , fese , and fete , which act as the hosts .
the main goals of such chemistry has been to both prepare new compounds and ultimately control the physical properties of the binary compounds in order to discover the underlying mechanism for superconductivity .
we believe the reader will recognize , however , that this chemistry has broader implications than superconductivity and could represent a new area for the preparation of two - dimensional ( 2d ) inorganic materials .
this year , 2016 , is a leap year since february has an extra day .
such an insertion of a day into the calendar is known as an intercalation , and chemistry has borrowed this term to analogously describe the expansion in a solid upon inserting a foreign species .
intercalation chemistry in inorganic materials has a long history,@xcite and its applications range from electrochemical energy storage@xcite to the sequestration of waste material in layered clays or sulfides.@xcite it is useful to briefly review the terminology in this field and some of the work done on past superconductors .
like closely related interstitial compounds,@xcite intercalation compounds are inorganic solids in which guest species such as ions or molecules occupy open crystallographic sites of the host . unlike interstitial compounds , however , intercalation hosts usually consist of structural elements such as chains and layers that are held together by weak forces such as those of the van der waals type .
the guest species , or _ intercalate _ can cause the unit cell of the host to expand upon insertion but it should not change the crystal structure significantly . furthermore , the intercalation and de - intercalation processes should be reversible and possible at temperatures lower than those typical of solid state reactions so that no major rearrangement of the host s structure occurs . given the constraints of the intercalation processes , two dimensional ( 2d ) materials are the most common hosts.@xcite materials with one - dimensional ( 1d ) motifs can facilitate exit and entry of guest species but also undergo major re - organization upon intercalation .
conversely , three - dimensional ( 3d ) materials retain their structural integrity but impede facile intercalation of guests .
classic inorganic solids in the 2d category include layered materials such as graphite,@xcite graphite oxide,@xcite and montmorillonite a silicate clay.@xcite if the host is redox active , then the intercalation process is either reductive or oxidative , and this special property has been explored in electrochemical devices such as rechargeable batteries.@xcite in order to raise some key questions from the literature on iron chalcogenides , we briefly discuss three other inorganic hosts that enhance their superconducting properties upon intercalation .
these include the layered metal dichalcogenides , layered cobaltates , and three - dimensional fullerene crystals .
these cases present to us the importance of charge doping , lewis acid - base chemistry , and structure - property relationships on superconductivity , and guide our understanding of iron chalcogenide intercalation chemistry .
@xmath10(nh@xmath11 ) where the nitrogen atoms ( blue ) are in a hexagonal arrangement between the sulfide sheets .
crystal structure from ref .
@xcite b. ) a representation of the orientation of the ammonia molecule so that the 3-fold axis is oriented parallel to the tas@xmath12 sheets .
reprinted with permission from ref .
copyright 2008 , aip publishing llc . ]
metal dichalcogenides are redox active hosts with a van der waals gap between the layers that allows facile intercalation.@xcite furthermore , they are semiconducting , which aids in the insertion and removal of cations by conducting charge - compensating carriers ( _ i.e. _ electrons and/or holes ) . in the early 1970 s , several groups demonstrated that upon intercalation , the superconducting critical temperature , or @xmath1 , could be raised in dichalcogenides such as tas@xmath12 and nbse@xmath12.@xcite in particular , the work of geballe et al .
demonstrated how a large array of organic and inorganic hosts could act as lewis bases , or electron donors , and therefore easily intercalate into the layered materials that act as lewis acids.@xcite the layered hosts can accept lewis bases such as ammonia , pyridine , or even amides with long aliphatic chains such as stearamide . since the metal dichalcogenides act as lewis acids , the @xmath1 s of the final adducts were related to relevant parameters such as interlayer spacing and the @xmath13 s of the lewis bases .
lerf has raised doubts , however , whether lewis acid - base chemistry truly drives the intercalation process in dichalcogenides.@xcite the chief objection from lerf concerns the findings from crystallographic studies showing that the orientation of the nh@xmath11 molecule within the layers is such that its 3-fold axis is oriented along the layers as opposed to perpendicular to them as shown in figure [ fig_nh3-tas2 ] for the case of tas@xmath12.@xcite similarly , the 2-fold axis of the pyridine ring is oriented parallel to the dichalcogenide layers.@xcite this would imply that the lone pair on the nitrogen atom is not oriented towards the layers and is therefore ineffective as a lewis base . besides the electron donating properties upon intercalation , structure - property relationships are also important .
the common structural motif is the 2d layer formed by infinite edge - sharing @xmath14 polyhedra where @xmath15 = transition metal and @xmath6 = s , se , and te .
polytypism , is common in these materials , since the coordination of the transition metal can be either octahedral or trigonal prismatic due to changes in the ion stacking sequence along the interlayer direction ( figure [ fig_nh3-tas2 ] ) .
however , only when the metal is in trigonal prismatic coordination has superconductivity been observed.@xcite in their studies of 50 different intercalates of various sizes for tas@xmath12 , which increased the interlayer spacing anywhere between 3 to 52 , gamble et al
. found the @xmath1 bears little correlation to interlayer spacing.@xcite instead , they attributed any enhancement in @xmath1 to the efficacy of charge transfer from the intercalate to the host .
alkali metals can also be intercalated into dichalcogenide hosts , either in a liquid ammonia solution or electrochemically , which allows for a more direct measure of electron doping than the lewis basicity of amines.@xcite furthermore , schllhorn,@xcite colombet,@xcite and others@xcite have argued that the intercalation of amines actually occurs through the oxidation of such species so that the true nature of the intercalates is the amine plus its conjugate acid ( _ e.g. _ nh@xmath11 and nh@xmath16 ) , which would also charge dope the dichalcogenide host .
the formation of cationic intercalates , in fact , has been suggested as the driving force for the intercalation of such species into the host.@xcite once the cationic form of the amines has been inserted , the other amines orient themselves in such a way to direct the lone pair of electrons on the nitrogen atoms towards the cations .
coo@xmath12@xmath10 1.3 h@xmath12o as a function of sodium intercalation .
the maximum @xmath1 is reached for nearly one - third doping and for the hydrated version .
figure is adapted from ref .
@xcite by permission from macmillan publishers ltd : nature , copyright 2003 . ]
a more recent example of intercalation chemistry than that of the dichalcogenides includes the layered na@xmath17coo@xmath12 materials.@xcite these cobalt oxides consists of coo@xmath18 octahedra that edge - share to form layers of the brucite - type structure .
na and li cations can fill in the interstitial positions in between and the shuttling of these cations has made the cobaltates an important cathode material for li - rechargeable batteries.@xcite in the sodium intercalated case , when the right amount of na@xmath3 is present along with water molecules , superconductivity is induced .
interestingly , the anhydrous case is non - superconducting , and the highest @xmath1 of 4.5 k is found for the stoichiometry of na@xmath19coo@xmath12 @xmath10 1.3 h@xmath12o ( see figure [ fig_na - coo2]).@xcite at values for @xmath20 greater than @xmath21 , anhydrous na@xmath17coo@xmath12 can exhibit various ordered states for its charge and spin degrees of freedom.@xcite these structures , much like the metal dichalcogenides , demonstrate that they can can accommodate more than one type of guest simultaneously , and that this can be advantageous for superconductivity .
another historical example is that of the intercalated fullerenes . just as graphite
can be easily intercalated,@xcite the molecular allotrope of carbon , c@xmath9 can also accommodate metal cations within its interstices . in the 1990 s
, researchers found that when fullerenes are charge - doped to become fullerides , they are tuned from semiconducting materials to metals and superconductors.@xcite the molecular nature of the crystal and the relatively weak interaction between the c@xmath9 ` hard spheres ' make intercalation possible not only for alkali metals but also molecular species such as ammonia.@xcite a notable difference in the intercalation chemistry of c@xmath9 and graphite is the former s preference for reductive intercalation whereas graphite can undergo both oxidative and reductive insertion.@xcite remarkably , the intercalated fullerenes such as k@xmath11c@xmath9 , rb@xmath11c@xmath9 , and rb@xmath12csc@xmath9 exhibit relatively high @xmath1 s , up to 32 k.@xcite .
the historical examples given above illustrate some of the key factors to consider for the iron - based superconductors .
these considerations include the role of the intercalate in charge doping into the density of states near the fermi level , distorting the relevant structural parameters , and changing the dimensional nature of the fermi surface .
nevertheless , all of these examples are classified as conventional superconductors whose underlying physics is satisfactorily described by bardeen - cooper - schrieffer ( bcs ) theory.@xcite therefore not all lessons learned from these studies will apply to the iron - based materials .
before discussing the intercalation chemistry of the iron based superconductors , it is instructive to describe the structure , bonding , and electronic properties of the host materials .
the fe@xmath6 ( @xmath6=s , se , te ) layers , like the transition metal dichalcogenides , are held by weak van der waals forces that make them susceptible to intercalation . in the case of the iron telluride , however , the van der waals gap is too small on account of the large anionic radius of te@xmath8 ( 2.21 ),@xcite and this hinders any significant intercalation chemistry .
therefore , this review is focused on the selenides and sulfides , but we do present the electronic and physical properties of fe@xmath22te ( where @xmath20 represents excess iron ) briefly in order to compare and contrast with those of fes and fese . with the anti - pbo type structure .
setting 2 for the origin is represented for this tetragonal space group .
the dashed line represents the unit cell , and @xmath23 is the so - called anion height . ]
although many binary iron sulfides and selenide phases are known , we are only interested in those that crystallize in the anti - pbo type structure .
as shown in figure [ fig_fech ] , the structure is described by the tetragonal space group @xmath24 ( no .
129 ) , which has two settings . in setting 1 ,
the iron atoms are located at the origin of the cell while in setting 2 the origin is located at the center of inversion symmetry with the iron atoms displaced ( @xmath25,@xmath26 , 0 ) from the origin.@xcite figure [ fig_fech ] shows setting 2 .
either way , the site symmetries of the atoms are such that the fe cations have @xmath27 site symmetry and the @xmath28 anions @xmath29 symmetry . furthermore , the @xmath30-glide normal to the @xmath31-axis leads to the adjacent layers being staggered by ( @xmath32 ) in the @xmath33-plane , which maximizes the distance between the @xmath28 anions . in the resulting two - dimensional square lattice of fe cations ,
each square is capped by a chalcogenide anion , either above or below , which leads to tetrahedral coordination for each iron cation .
the only crystallographic parameters that can be manipulated include the lattice constants and the @xmath6 anion @xmath23-parameter , also know as the anion height from the iron square lattice .
these two parameters in addition to the lattice constants are therefore enough to describe important properties such as fe se bond distances , fe fe distances , which influence the @xmath34-orbital dispersion bands , and the @xmath6fe@xmath6 bond angle .
as shown in figure [ fig_structrends ] , the anion height and the variance of the tetrahedral bond angle both correlate with @xmath1 across the various iron - based superconductor families.@xcite structurally , the layered iron chalcogenides provide a nice contrast to the transition metal dichalcogenides . the former is composed of a square planar lattice of @xmath34-block metals whereas the latter is composed of a hexagonal array . in the iron - based chalcogenides ,
2d layers consist of edge - sharing tetrahedra ( figure [ fig_fech ] ) whereas in the dichalcogenides @xmath14 trigonal prisms or octahedra complete the layers .
these two structural distinctions lead to differences in both the local electronic properties determined by crystal field splitting considerations and to the long - range electronic properties determined by the band structures . with the anion height parameter for various members of the iron arsenide and chalcogenide superconductors .
reprinted with permission from ref .
copyrighted by the physical society of japan .
bottom : the correlation of @xmath1 with the tetrahedral bond angle @xmath35 for the iron arsenide family of superconductors . reprinted with permission from ref .
copyright iop publishing . reproduced with permission .
all rights reserved . ]
one of the most studied iron - based superconductors is iron selenide . in its anti - pbo structure
it is known as @xmath0-fese .
this phase exists in a narrow window of stability in the iron - selenium binary phase diagram ( 300 k@xmath36 450 k),@xcite as shown in figure [ fig_phasedia - fese ] , and its preparation has mostly consisted of powders through direct reaction of the constituent elements .
polycrystalline fese , however , can be prepared through hydrothermal routes , and care must be taken to prevent oxidation , which is detrimental to the superconducting properties.@xcite since this phase does not melt congruently , single crystal growth has been challenging .
several groups have found chemical vapor transport and salt flux methods , however , as successful techniques for the preparation of high quality single crystals.@xcite when near stoichiometric , fese superconducts below 8 k.@xcite interestingly , under high pressure the @xmath1 increases up to 37 k.@xcite even higher @xmath1 s have been observed in fese in isolated single layers . through _ in situ
_ electrical resistivity measurements during photoelectron spectroscopy measurements , @xmath1 s have ranged from 65 k to 100 k in single layered fese.@xcite this is the record for any iron - based superconductor , and it raises the possibility that materials chemists and physicists could develop a new family of high-@xmath1 superconductors . above @xmath1 fese undergoes a crystallographic phase transition that may provide clues to its superconducting pairing mechanism . through various diffraction
studies,@xcite a transition from @xmath24 to orthorhombic @xmath37 at 90 k in superconducting fese was observed .
this lowering of symmetry is accomplished through the breaking of the four - fold axis and the @xmath30-glide plane normal to the @xmath31-direction . however , for the non - superconducting fe@xmath38se , which contains extra interstitial iron , the symmetry lowering is not observed , and the driving mechanism lifting the degeneracy of the @xmath39-plane was proposed to not be magnetically driven.@xcite more recent diffraction studies by pachmayr et al.@xcite have found that fese prepared by hydrothermal conditions as opposed to traditional solid state techniques actually leads to a triclinic system and a lower temperature of transition near 60 k. systematic diffraction studies including neutrons at lower temperatures of various fese single crystal samples would help shed further light on how composition drives the symmetry lowering transition .
shear - modulus@xcite and angle resolved photoemission spectroscopy ( arpes ) studies have proposed that the orthorhombic distortion leads to a so - called nematic phase where the stripes of elongation are driven by an orbital - ordering of the @xmath34-states.@xcite however , this topic remains unsettled with new results from neutron studies@xcite and theoretical modeling@xcite that report the possible importance of spin fluctuations in fese .
-fese shown in blue .
reprinted figure with permission from ref .
copyright 2009 by the american physical society . ] in contrast to @xmath0-fese , the sulfide analogue does not exist in the thermodynamic phase diagram of iron sulfides . within the binary phase diagram of the iron sulfur system
, there are approximately 7 different compositions with 10 different crystallographic structures.@xcite many of the phases are well known minerals such as pyrite ( fes@xmath12 ) or pyrrhotite ( fe@xmath40s )
. therefore tetragonal fes with the anti - pbo type structure , also known as mackinawite , is a metastable phase . in the early crystallographic studies , the diffraction data set was not of high quality enough due to sample broadening , and the @xmath23-parameter of the s@xmath8 anion
was fixed to @xmath41.@xcite subsequent studies by lennie et al . found @xmath23 to be larger than 0.25 in order for the fes@xmath42 tetrahedron to have more regular tetrahedral angles.@xcite the various thermal studies by lennie et al . found that while well crystallized mackinawite eventually decomposes to pyrotite near 530 - 545 k,@xcite powder samples can first convert to greigite fe@xmath11s@xmath42 at a much lower temperature of 373 k.@xcite therefore
, it is likely that the larger particle sizes and therefore samples with smaller surface areas have a higher barrier towards conversion of tetragonal fes to one of its thermodynamically stable phases .
fes is typically prepared by solution routes either under ambient conditions@xcite or hydrothermally.@xcite as a polycrystalline material , mackinawite can be prepared by oxidation of iron metal in an acetic acid / acetate buffer solution .
afterwards , an aqueous solution of na@xmath12s hydrate is added to the buffer leading to the rapid precipitation of tetragonal fes.@xcite according to the pourbaix diagram of iron , under basic conditions , fe@xmath43 species predominate whereas fe@xmath4 predominate at lower phs.@xcite therefore , in the initial low - ph solution , the iron is oxidized to evolve fe@xmath4 species , and the buffer serves to minimize the amount of fe@xmath43 species upon adding the sodium sulfide .
lennie et al . have argued that mackinawite is kinetically stabilized and transforms to other phases only through oxidation or further sulfurization.@xcite studies on the physical properties of mackinawite fes have been inconsistent with some studies reporting a semiconducting ground state@xcite and others superconductivity.@xcite nanocrystalline fes , prepared by the crystallization of an amorphous black solid through a solvothermal method , was found to be semiconducting and ferrimagnetic.@xcite likewise denholme et al .
found their mackinawite to be semiconducting@xcite but upon the application of pressure , the materials become more metallic.@xcite denholme et al .
suggest that mackinawite may indeed be metallic , and that the semiconducting behavior arises from grain boundaries in the polycrystalline samples .
ferrimagnetism was also observed in the magnetization vs. applied field curves , and the magnetic susceptibility vs. temperature curves indicate that the nel temperature is above room temperature .
a much earlier study , however , had found no long range magnetic ordering in mackinawite .
bertaut performed neutron diffraction studies on powder samples of mackinawite.@xcite in the powder pattern at 10 k , antiferromagnetic bragg peaks were not observed .
bertaut concluded that fes is not antiferromagnetic and can not therefore be ferrimagnetic .
furthermore , electronic structure calculations had predicted that , much like its heavier analogues fese and fete , mackinawite should be a metal or semimetal.@xcite the observation by lennie et al . that the surface of mackinawite first converts to fe@xmath11s@xmath42@xcite
could explain the ferrimagnetic behavior in magnetization measurements since fe@xmath11s@xmath42 , which has the spinel - type structure , is a ferrimagnetic semiconductor.@xcite very recently , lai et al .
found mackinawite fes to be metallic down to 5 k , below which it becomes superconducting.@xcite lai et al . prepared a high quality sample of tetragonal fes by oxidizing iron metal under hydrothermal conditions in the presence of na@xmath12s hydrate ( figure [ fig_tem - fes ] ) .
likewise , borg et al . demonstrated that single crystals of k@xmath17fe@xmath44s@xmath12 could be de - intercalated hydrothermally to prepare single crystals of fes.@xcite the key to obtaining superconducting crystals was to maintain the solution under basic conditions and to provide enough iron powder during the hydrothermal synthesis in order to fill in the vacancies in the k@xmath17fe@xmath44s@xmath12 samples .
the single crystals afforded the true ground state properties of mackinawite , which include highly anisotropic superconducting and normal state properties .
this high anisotropy is indicative of the 2d electronic structure of fes . above @xmath1 , fes is truly a metal with weak paul paramagnetism.@xcite while fe@xmath22te is an interesting parent phase ( _ i.e . not superconducting until doped_),@xcite we have not seen any literature on its intercalation chemistry .
the difficulty may arise for several reasons including the smaller van der waals gap due to the larger te@xmath8 anionic radius with respect to its lighter congeners .
furthermore , intercalation often occurs under highly basic conditions , and te is more easily oxidized under these conditions than se or s since it is more electropositive .
de - intercalation chemistry on fe@xmath22te and fe@xmath22te@xmath45se@xmath19 , however , has been demonstrated in order to remove the interstitial iron , @xmath20,@xcite which has large implications for the magnetic , transport , and superconducting properties.@xcite .
therefore , the iron telluride system may still be an interesting host structure to explore for possible intercalation , ion exchange , and de - intercalation chemistry .
finally , it is important to discuss the electronic structure of the host structures .
the square fe sublattice has a strong 2d character in its electronic structure .
the fe@xmath6 layers are metallic in nature , and the @xmath34-states of the fe cations form bands near the fermi level . unlike the previous intercalated superconductors , the iron - based materials are multi - band superconductors .
the 2d character is clearly observed in the cylindrical nature of the fermi surfaces as found by density functional theory ( dft ) calculations ( figure [ fig_fechfermisurf]).@xcite in this respect , the chalcogenides do have some similarities to the iron pnictides , where nesting between the electron and hole pockets leads to a spin density wave instability and crystallographic phase transitions.@xcite upon charge doping , the sdw vanishes and the pnictides become superconducting.@xcite while sdw order has been found in the fe@xmath22te phases , no such order has been observed in fes and fese .
the fermi surface of fes and fese show several separate sets of bands : one set is zone - centered and are hole carrying , and the other set is corner - centered and electron carrying.@xcite the dispersion curves reveal that these bands consist of the various fe @xmath34-orbitals while the @xmath6 @xmath46-orbitals are buried deep in energy . according to a recent arpes study of fese , the bands around the zone center consist largely of the @xmath47 and @xmath48 orbitals , which are degenerate in the tetragonal phase , but split upon the orthorhombic distortion near 90 k.@xcite the electron pocket meanwhile is dominated by the @xmath49 orbital . from their arpes study on high quality crystals , watson et al .
conclude that orbital - charge ordering of the @xmath47 and @xmath48 bands drive the crystallographic phase transition near 90 k in fese.@xcite a simplistic way to understand the band structure from a chemical bonding perspective is to consider the ligand field splitting for the fese@xmath42 tetrahedra . for tetrahedral coordination ,
the doubly degenerate @xmath50 states consist of the @xmath51 and @xmath52 orbitals .
the @xmath50-states are lower in energy than the triply - degenerate @xmath53-states , which consist of the @xmath47 , @xmath48 , and @xmath49 orbitals . in the case of the layered fe@xmath6 materials , rather than being antibonding the @xmath53-states are mostly nonbonding .
the local geometry around the iron cations , however is not ideally tetrahedral since the site symmetry of the cations is @xmath27 ( or @xmath54 ) and therefore lower than @xmath55 ( or @xmath56 ) .
this site symmetry lifts the degeneracy of the @xmath53-manifold , so that the @xmath47 , @xmath48 orbitals form their own degenerate pair .
perhaps not coincidentally , a reliable structural descriptor for the @xmath1 in iron - based superconductors is the closeness of the tetrahedral bond angles to the ideal 109.5@xmath57 , with the highest observed for the arsenide family being @xmath1 ( figure [ fig_structrends]b ) .
fe@xmath44se@xmath12 where @xmath58 can be various species including alkali metal cations or moieties such as nh@xmath11 and nh@xmath59 .
the structure is also known as the thcr@xmath12si@xmath12-type structure . ] before discussing the first attempts at intercalating fese and fes , we briefly mention that compounds of the type @xmath60fe@xmath44se@xmath12 had been prepared before through solid state techniques ( where @xmath61 = alkali metal or tl@xmath62 cations).@xcite these materials exhibit a significant increase in the @xmath1 up to approximately 30 k. since single crystals can readily be grown from congruent melts,@xcite we do not consider their preparation as intercalation chemistry .
we briefly describe them , however , since their crystal structures are related to the intercalated iron chalcogenides and because their shortcomings highlight the need for intercalation chemistry to properly study the superconducting properties of fese .
in contrast to the host structure , the alkali metal intercalated phases crystallize into the thcr@xmath12si@xmath12-type structure with a body - centered tetragonal cell and space group @xmath63 as shown in figure [ fig_afe2ch2 ] .
therefore the relationship between the two adjacent fe@xmath6 layers is different from that of the host . in the latter ,
the layers are staggered with respect to one another by the @xmath30-glide plane ( @xmath64 ) .
this is not the case in the body centered cell , which is missing the @xmath30-glide , and the @xmath28 anions are located directly over one another .
presumably , ionic forces now dominate the bonding between the layers as intercalation of the @xmath61 cations reduces the fe@xmath6 to become anionic ( fe@xmath6)@xmath65 layers . a serious drawback for these ternary selenides , however , is that the @xmath60fe@xmath44se@xmath12 superconducting phase coexist with an insulating phase,@xcite which was later found to be near the stoichiometry of @xmath66fe@xmath42se@xmath67.@xcite since the superconducting and antiferromagnetic phases are structurally related , they often coexist when prepared by congruently melting the substituent elements .
shoemaker et al
. demonstrated through their careful diffraction studies that the minority phase in the samples is superconducting,@xcite and several studies have shown that the amount of @xmath20 leading to superconductivity ranges from 0.3 to 1.0.@xcite intercalation at low temperatures into pristine fese layers therefore offer an opportunity to enhance superconductivity and avoid the iron vacancy - populated phase . while several groups have proposed that the stoichiometry of the superconducting phase to be close to @xmath60fe@xmath12se@xmath12,@xcite ding et al .
formulated the superconducting phase from their careful electron microscopy studies as approximately k@xmath12fe@xmath68se@xmath69.@xcite therefore , it is the particular ordering of vacancies in @xmath66fe@xmath42se@xmath67 that is thought to be detrimental to superconductivity , favoring instead long - range antiferromagnetic ordering at 559 k.@xcite svitylk et al
. further demonstrated that vacancy ordering in itself may not be detrimental by finding superconductivity in the rb@xmath17fe@xmath70se@xmath12 system , which showed some iron vacancies.@xcite interestingly , the magnetic moment per iron cation of 3.31 @xmath71 in k@xmath12fe@xmath42se@xmath67 suggests that the @xmath72 cations in the tetrahedral crystal field may have an orbital contribution in addition to spin.@xcite ( nd@xmath12)@xmath73(nd@xmath11)@xmath74fe@xmath12se@xmath12 where the li cations reside mostly between the se@xmath8 anions and the nd@xmath11 and nd@xmath59 moieties at the cell center .
reprinted from ref .
@xcite by permission from macmillan publishers ltd : nature , copyright 2012 . ] in a similar fashion to the intercalation of alkali metals into dichalcogenides , fese was reduced by an alkali metal - liquid ammonia solution , which contains solvated electrons ( hence its blue color).@xcite through this ammonia solution alkali and alkaline earth metals can be inserted into the fese host and the @xmath1 raised to 43(1 ) k.@xcite burrard - lucas et al .
have definitively demonstrated through neutron diffraction studies that this technique also intercalates ammonia and amide ( nh@xmath59 ) molecules along with li@xmath3 cations in the interstitial positions to form li@xmath75(nh@xmath12)@xmath76(nh@xmath11)@xmath70fe@xmath12se@xmath12 . unlike the typical thcr@xmath12si@xmath12-type structure , however , the li cations are predominately in the @xmath77 wyckoff position , which is not at the unit cell s body center but between the se@xmath8 anions as shown in figure [ fig_lind3-fe2se2 ] .
using deuterated ammonia for the neutron experiments , burrard - lucas found the nd@xmath11 and nd@xmath59 moieties to be located at the cell center , and the deuterium atoms disordered at the @xmath78 site.@xcite than the one with the single layer of nh@xmath11 , nh@xmath59 , and li@xmath3 species .
reprinted with permission from ref .
copyright 2014 american chemical society . ] in a follow up to their neutron diffraction study of li@xmath75(nh@xmath12)@xmath76(nh@xmath11)@xmath70fe@xmath12se@xmath12 , sedlmaier et al . found that an intermediate phase exists with a double layer of ammonia and amide molecules.@xcite their _ in situ _ diffraction experiment revealed that this double layer structure actually retains the primitive tetragonal cell of the parent host , but expanded as shown in figure [ fig_linh3-fe2se2 ] . by ammoniating li@xmath75(nh@xmath12)@xmath79(nh@xmath11)@xmath74fe@xmath12se@xmath12 , sedlmaier et al .
were also able to prepare li@xmath75(nh@xmath80)@xmath81fe@xmath12se@xmath12 , which lowered the @xmath1 from 43(1 ) k to 37(1 ) k.@xcite since this phase was identical to the intermediate phase found in the _ in situ _ studies , the intercalation of neutral species such as nh@xmath11 into these materials is reversible . for similar compounds , shylin et al
. demonstrated that the li cations do indeed charge dope the fese hosts through @xmath82fe - mssbauer spectroscopic studies,@xcite therefore confirming that the nh@xmath59 molecules do not completely charge compensate the li@xmath3 cations .
regime exists for samples with the stoichiometry of k@xmath19(nh@xmath11)@xmath76fe@xmath12se@xmath12 , the lowest-@xmath1 is for samples with stoichiometry of k@xmath75(nh@xmath11)@xmath76fe@xmath12se@xmath12 , and the middle regime is an unknown phase .
b. ) a clear difference in the @xmath31 lattice constant is observed , which correlates with @xmath1 and amount of charge doping into the fese layers .
reprinted with permission from ref .
copyright 2013 american chemical society . ]
several groups soon demonstrated that a host of different electropositive metals besides lithium such as potassium , rare earths , and alkaline earths could be intercalated into the fese sheets through the liquid ammonia method.@xcite ying et al .
found that the amount of k@xmath3 cations inserted was more important for affecting @xmath1 than the amount of ammonia intercalated.@xcite in their phase diagram , shown in figure [ fig_kfe2se2nh3 ] , the superconducting regime does not have a dome - like appearance but rather resembles three flat plateaus .
essentially , the highest @xmath1 in this system ( @xmath83 k ) is achieved for a stoichiometry of k@xmath19(nh@xmath11)@xmath84fe@xmath12se@xmath12 and the lower @xmath1 ( 30 k ) , which is coincidentally similar to the one with just alkali species , is found for the composition of k@xmath75(nh@xmath11)@xmath85fe@xmath12se@xmath12 .
an unidentified phase superconducts below 36 k , which is close to the @xmath1 found by sedlmaier et al . in their intermediate phase with the primitive cell and a double layer of ammonia molecules.@xcite we can conclude from their study , that there is an optimal amount of doping into the fese sheets , which is close to 0.15 electrons per iron cation , and that by overdoping , the @xmath1 is driven down .
it is not clear from this study , however , whether any amide species also intercalates into the fese as found for the lithium intercalated compounds by the group of clarke.@xcite through careful _ in situ _ x - ray diffraction studies , yusenko et al . found even more intermediate phases during the intercalation of cations via liquid ammonia.@xcite initially , a fast intercalation occurs at 200 k with a characteristic time scale , @xmath86 , of 2 - 3 min .
the phase prepared by the fast intercalation includes two layers of ba@xmath4 cations with ammonia molecules in between .
the slower process is the rearrangement of the guest species , @xmath86 = 70 min , to form a phase similar to the one found by burrard - lucas et al .
for li,@xcite but with more ammonia positions . upon warming ,
the body - centered phases are not stable and decompose to the primitive setting , and upon applying vacuum all the nh@xmath11 can be removed to reach a final composition of ba@xmath87fe@xmath12se@xmath12 , which exhibited a @xmath1 of 34 k. overall , the highest @xmath1 found was close to 39 k for the ammoniated body - centered phase ba@xmath87(nh@xmath11)@xmath88fe@xmath12se@xmath12 . the entire sequence of powder patterns and the corresponding intercalated phases can be found in figure [ fig_banh3-fe2se2 ] . just as fese can have its @xmath1 enhanced by the application of pressure , izumi et al .
demonstrated that pressure can also enhance the @xmath1 of the intercalated selenides.@xcite after preparing several samples with doped alkali metals , izumi found a correlation between the tetragonality ( @xmath89 ratio ) of the unit cell and the @xmath1 .
in general , the higher the tetragonality the higher the @xmath1 , although the affect seems to plateau beyond a certain @xmath89 ratio . for their cs - nh@xmath11 intercalated samples ,
the application of pressure initially drives down the @xmath1 , eventually wiping out all superconductivity . above the threshold of 14 gpa ,
however , superconductivity re - emerges and @xmath1 reaches to a high of 49 k until it drops to 0 above 21 gpa .
the resulting phase diagram has a double dome appearance for the superconducting regime as shown in figure [ fig_csfese ] , and izumi et al .
ascribe this phenomena to the proximity of a quantum critical point.@xcite the authors do not observe any crystallographic phase transitions upon applying pressure .
-rich or -poor structures .
b. ) a schematic of the structures for the xrd patterns of a. ) , although the orientation of the ammonia molecule was not solved from the diffraction data .
c. ) the squid magnetometry data for the na - intercalated fese , and d. ) the phase diagram for the na - intercalated fese@xmath90s@xmath23 hosts . reprinted by permission from ref .
@xcite , macmillan publishers ltd : nature communication , copyright 2014 . ]
while no study has been published on the intercalation of amines into fes or fete , some groups have attempted to intercalate mixed anion systems .
for example , while the highest @xmath1 observed for the optimal stoichiometry of fese@xmath91te@xmath91 is 14 k , sakai et al
. demonstrated through intercalation of li , na , and ca cations via the liquid ammonia technique that the @xmath1 could be raised to 26 k , 22 k , and 17 k , respectively.@xcite .
similarly , guo et al .
intercalated fese@xmath90s@xmath92 with na cations and found that the @xmath1 could also be enhanced from that of the of host structure ( @xmath93 k).@xcite as more sulfide is introduce into the lattice , the @xmath1 is diminished and the phase no longer superconducts past a certain amount of s@xmath8 substitution . the phase diagram for their phases
is presented in figure [ fig_nafe2ses ] .
similar to the conclusions reached by sedlmaier et al . on li - intercalation,@xcite guo et al . found a clear relationship between the structure - type and the maximum @xmath1 ( figure [ fig_nafe2ses]).@xcite to understand the various findings from the experimental work on the intercalated iron selenides , guterding performed electronic structure calculations to understand the role of the doping cations and ammonia spacers on the superconducting properties.@xcite focusing specifically on the li@xmath91(nh@xmath12)@xmath94(nh@xmath11)@xmath95fe@xmath12se@xmath12 system , they found that the major determining factor on the characteristics of the fermi surface was the amount of electron doping .
as shown in figure [ fig_linh3-fesefermisurf ] , when the amount of amide present is zero ( @xmath96 ) , the li cations donate approximately 0.25 electrons per fe cation .
the electron pockets at the corners expand and the hole pockets at the center decrease in size ( figure [ fig_linh3-fesefermisurf]c ) .
when amide is added so that the charge of the li is completely compensated , the opposite occurs and a new hole pocket is created figure [ fig_linh3-fesefermisurf]b ) .
what is also evident from these calculations is that the bands are dominated by the @xmath49 , @xmath47 , and @xmath48 orbitals .
guterding et al .
conclude that the amount of spacer molecules such as ammonia is not the determining factor for @xmath1 , but it can aid in making the fermi surface more 2d ; after that has been accomplished , the amount of electron doping will then enhance the superconducting properties . to conclude this section ,
intercalation of metal cations in liquid ammonia in fese hosts seems to be the best manner to prepare higher @xmath1 ( @xmath83 k ) phases that are single phase and therefore free of the insulating antiferromagnetic iron selenides .
the role of the ammonia is not just as a solvent that allows the fese hosts to absorb the solvated electrons , but also as a spacer molecule that facilitate the intercalation process .
furthermore , the ammonia can ionize to form amide in solution with li@xmath3 cations . so far
, no one has reported the formation of the conjugate acid nh@xmath16 in these intercalated compounds . during intercalation
a series of intermediates are formed as shown by the various _ in situ _ diffraction studies , and they can be identified on the basis of whether they exhibit primitive or body - centered lattice symmetry .
the spacer aids in further enhancing the 2d nature of the fermi surface , but the electron doping seems to be the major determining factor of the maximum @xmath1 observed .
just as in the transition metal dichalcogenides , the community soon recognized that other amines besides ammonia could be intercalated into the iron chalcogenides .
the first such report was performed by krzton - maziopa et al . where they used anhydrous pyridine as the solvent for a variety of alkali metals including li , na , k , and rb.@xcite by reacting with pure @xmath97fese at 40 @xmath57c , krzton - maziopa et al .
managed to expand the @xmath31-axis and enhance the @xmath1 up to 45 k , close to the observed values for the ammonia intercalated selenides .
as shown by powder xrd , the fese sheets are indeed intercalated , but the lattice constants however are not that different from those in the ammonia - intercalated samples .
for example , the li - pyridine intercalated sample had a @xmath31 parameter of 16.0549 ,@xcite which is close to the values of 16.1795(6 ) and16.4820(9 ) observed by burrard - lucas et al .
for the li - nh@xmath11 intercalated samples.@xcite furthermore , the k - intercalated samples also had a smaller @xmath31-parameter , which would be unexpected for the larger pyridine ring intercalated with the larger potassium cation .
krzton - maziopa et al . also observed licn as a byproduct from the reaction , which could indicate that the pyridine rings decomposed and therefore the nature of the amine intercalate could be something other than pyridine.@xcite the work by the group of noji et al .
showed that linear diamines could also be utilized as solvents and intercalates for fese.@xcite starting with the classic coordination ligand , ethylenediamine ( c@xmath12h@xmath69n@xmath12 ) , hatakeda et al .
intercalated this ligand along with li to raise @xmath1 up to 45 k. the intercalation reaction seems to be sluggish with a total of 7 days at 45 @xmath57c followed by post - annealing at 150 @xmath57c and 200 @xmath57c for several more days under vacuum.@xcite assuming that the compound crystallized in the body - centered cell , the authors indexed the lowest angle peak as the ( 002 ) reflection and therefore the interplanar distance to be close to 10.37 ( figure [ fig_aeda - fese ] ) . by expanding to even larger chains such as hexamethylenediamine ( c@xmath18h@xmath98n@xmath12 ) , hosono et al .
increased this interplanar distances to a record 16.225(5 ) .@xcite this hyper - expansion of the fese host , however , does not seem to have increased @xmath1 but rather lowered it to 38 k , which is in line from the previous studies of metal - ammonia intercalation whereby charge doping is more important than the size of the spacer .
in addition to lithium , other metals can be intercalated with ethylenediamine such as sodium .
figure [ fig_aeda - fese ] shows the xrd powder patterns of the li and na intercalated compounds along with the c@xmath12h@xmath69n@xmath12 spacers that lead to a body - centered cell and an interplanar distance close to 10.37 regardless of the choice of alkali metal.@xcite although the authors claim a @xmath99 as high as 57 k from electrical resistivity measurements , the magnetic susceptibility data shows that the true @xmath1 is closer to 45 k , which is more in line with other intercalated fese superconductors.@xcite a systematic study of various linear diamines with either na@xmath3 or sr@xmath4 cations was carried out by hayashi et al . to understand the role of charge and spacer length.@xcite the linear diamines include ( h@xmath12n)-c@xmath100h@xmath101(nh@xmath12 ) where @xmath102 0 , 2 , 3 , and 6 , and hayashi et al
. found that the interplanar distance could indeed be systematically varied between 8.95 and 11.16 for the sodium series of compounds .
no correlation , however , was found between the length of the diamine chain and @xmath1 . instead , the correlation was strictly between charge doping and @xmath1 . for materials with the sodium content varying between 0.80 and 0.93 ( per fe@xmath12se@xmath12 unit ) , the @xmath1 was between 41 k and 46 k , whereas for the strontium content , the @xmath1 was between 34 k and 38 k. typically the amount of sr varied between 0.23 to 0.28 ( per fe@xmath12se@xmath12 unit).@xcite overall , the recent work of intercalating organic amines with alkali and alkaline earth metals has demonstrated that the maximum @xmath1 is not enhanced beyond that of ammonia with alkali and alkaline earth metals .
the spacer , however does seem to have an effect below a threshold value for the interplanar distance .
as shown in figure [ fig_tcs - fese ] for a variety of compounds from fese to those with large interlayer spacers , the @xmath1 plateaus above a distance of approximately 8 .
it would be interesting for future work to understand any advantages in utilizing organic amines over ammonia such as thermal stability or finer control over electron doping .
however , caution must be exercised since working ethylenediamine can cause fragmentation of the fese layers such as fe@xmath11se@xmath42(en)@xmath12.@xcite and the interlayer spacing , @xmath34 , by the intercalation of various species including alkali metal cations , ammonia , and organic amines .
reprinted from ref .
@xcite , copyright 2014 , with permission from elsevier . ]
an interesting development in the intercalation chemistry of fese was the demonstration by lu et al . that such reactions could occur in aqueous solutions as opposed to ammoniacal ones that were strictly anhydrous.@xcite under hydrothermal conditions , lu et al .
added a selenide source in the form of selenourea to a strongly basic solution of lioh along with iron metal .
the resulting compound was formulated as lifeo@xmath12fe@xmath12se@xmath12 , which was structurally related to the lnofeas superconductors and the zrcusias - type structure ( figure [ fig_liohfese ] ) .
an electronic structure calculations study by heil et al . predicted that the ground state of this oxide - intercalated fese should be that of an antiferromagnet due to the effect of the @xmath103 states ( from fe@xmath43 ) in the oxide layer.@xcite however , this work did not yet have the full experimental details of the chemical composition and missed the protons present in the oxide layer , which would therefore call into question whether fe@xmath43 is really present in this layer as opposed to fe@xmath4 . soon after the work of lu et al .
, pachmayr et al . were able to prepare small single crystals from hydrothermal methods ( figure [ fig_liohfese - sem ] ) and found residual electron density corresponding to hydrogen from their diffraction measurements.@xcite therefore , pachmayr et al .
formulated the intercalate to actually be a hydroxide instead of oxide , and their single crystal x - ray diffraction results showed that the composition is close to ( li@xmath74fe@xmath73oh)(fe@xmath104li@xmath105se ) .
the hydroxide intercalate has a structure very similar to lioh , which itself crystallizes in the anti - pbo - type structure .
furthermore , antisite substitution occurs so that some iron cations substitute onto the li site , and _
vice versa_. pachmayr et al .
also established that the nature of the iron in the hydroxide layer is fe@xmath4 and not fe@xmath43 based on the fe
o distance of 201.6 pm .
the fe@xmath4 cations act to both electron dope into the fese host and to flatten the ( li / fe)o@xmath42 tetrahedra since fe@xmath4 prefers more square planar coordination.@xcite
cations are doped into the li sites , which charge dopes the fese hosts .
the lioh intercalate has the anti - pbo - type structure . ]
fe@xmath73oh)fe@xmath104li@xmath105se single crystals prepared by hydrothermal synthesis .
b. ) the electrical resistivity and magnetic susceptibility of ( li@xmath74fe@xmath73oh)fe@xmath104li@xmath105se , and the low temperature signal below 20 k represents a ferromagnetic response that coexists with superconductivity . reprinted with permission from ref . @xcite .
copyright 2015 wiley - vch verlag gmbh and co. kgaa , weinheim . ]
interestingly , due to the iron substitution onto the hydroxide layer , pachmayr et al . found the compound to express both ferromagnetism and superconductivity at base temperature , which led to an interesting phenomena by which spontaneous vortex lattice forms in this type - ii superconductor.@xcite as shown in figure [ fig_liohfese - sem ] , the zero - field cooled curve in the magnetic susceptibility shows and a ferromagnetic signal even in the diamagnetic regime . through nuclear magnetic resonance ( nmr )
studies of this compound , lu et al . found the hydroxide layer to actually exhibit antiferromagnetism instead of ferromagnetism.@xcite a small angle neutron scattering ( sans ) study found that under a small applied magnetic field , a vortex lattice was found in ( li@xmath40fe@xmath17od)fese in the pattern with a characteristic length not unlike in other superconductors.@xcite lynn et al
. concluded , however , that the vortex forms only under an applied magnetic field and is not spontaneous .
fe@xmath44se@xmath12 to ( lioh)fese .
b. ) the resulting change in the powder patterns from a body centered structure to the primitive cell . reprinted figure with permission from ref .
copyright 2015 by the american physical society . ] several studies on the phase diagram of the ( li@xmath40fe@xmath17oh)fese then followed to understand how the structural and chemical parameters controlled @xmath1 .
sun et al . found that instead of preparing the hydroxide samples _ in situ _ under hydrothermal conditions , that fese could act as the source for the preparation of ( li@xmath40fe@xmath17oh)fe@xmath70se where ( @xmath106 and @xmath107).@xcite the samples prepared in this manner repeatedly led to lower @xmath1 s , however , so sun et al .
performed post - synthetic modification of these powder samples by treating them with an ammoniacal solution of li metal .
this reductive post - synthetic modification led to maximal @xmath1 of 45 k , and sun et al .
explain this result as filling in iron vacancies in the fese layer.@xcite indeed , the technique of post - synthetic modification is a form of intercalation chemistry .
the large spread in iron deficiency found in their samples led sun et al . to quantitatively correlate iron occupancy in the selenide layer to superconductivity and to conclude that full iron occupancy in the selenide layer along with iron with an oxidation state below + 2 are required for superconductivity .
the area of soft chemistry can be further expanded beyond powder to include the preparation of single crystal samples of ( li@xmath40fe@xmath17oh)fese .
dong et al . found that starting with k@xmath17fe@xmath12se@xmath12 single crystals prepared at high temperatures , one could perform a cation exchange under hydrothermal conditions to prepare liohfese single crystals.@xcite such exchange and the resulting x - ray diffraction patterns are shown in figure [ fig_ionex - fese ] both the work of dong et al . and sun et al .
demonstrate that the iron selenides exhibit considerable soft chemistry that allows for facile preparation of superconductors with optimal properties .
zhou et al . were able to construct two phase diagrams , one for ( lioh)fese and the deuterated ( liod)fese from single crystal and powder samples .
no isotope effect was observed from the deuterium substitution . in their phase diagrams ,
zhou et al .
correlate the @xmath1 to the tetragonality of the unit cell ( @xmath89 ratio ) , which might be related to the overall oxidation state of the iron in the fese layer and therefore to the amount of charge doping from the hydroxide layer.@xcite dong et al .
also correlated @xmath1 to the lattice parameters but added an antiferromagnetic phase for samples with very small @xmath31 lattice constant , which they referred to as the parent compound to the superconductor.@xcite this antiferromagnetic ordering observed by dong et al . in their magnetization studies , however , was disputed by zhou et al . since they demonstrated that under hydrothermal conditions various ferrimagnetic iron oxides such as magnetite can be prepared .
furthermore , through their neutron diffraction studies of deuterated samples , zhou et al
. found no evidence for long range antiferromagnetic ordering either in superconducting or non - superconducting samples.@xcite this result from zhou et al
. would be in line with the vast literature on iron selenides , where a parent compound with long - range antiferromagnetic ordering has yet to be found.@xcite the insertion of hydroxide layers into other hosts has been accomplished as well , notably mackinawite fes . while the host fes has very recently been shown to be superconducting with a @xmath1 close to 5 k , pachmayr found the lioh - intercalated phase of fes to be non - superconducting and instead displays only weak ferromagnetism.@xcite likewise , lu et al .
found that in hosts with mixed anions such as fese@xmath90s@xmath92 , lithium hydroxide intercalation could be achieved but that the @xmath1 drops off dramatically as the amount of sulfide is introduced into the lattice.@xcite this could be due to the general instability of mackinawite fes , which might cause sufficient iron vacancies to form upon intercalation so that they completely disrupt the conducting properties .
the iron chalcogenides , and in particular fese , have been proven to be versatile hosts for intercalation chemistry .
furthermore , intercalation chemistry affords superconducting samples that are phase pure and therefore reveal the true ground state properties .
these low - temperature synthetic techniques are superior to that of simply melting the constituent elements ( alkali metal , iron , and selenium ) , since the latter leads to phase separation and the inclusion of an antiferromagnetic , insulating phase .
the intercalation chemistry of fese therefore demonstrates that solid state chemists will continue to play a role in furthering our understanding of superconducting materials .
the major parameters that seem to be important across the entire series of iron chalcogenides include structure and charge doping .
as such , we have gathered the relevant parameters across various intercalated selenides in table [ parameters ] for comparison .
although there have been many efforts to correlate @xmath1 with charge doping there are many inconsistencies apparent in table [ parameters ] , and this may be on account of the difficulty on determining whether the amines intercalated are neutral or charged .
furthermore , due to the nature of the disordered intercalates and sometimes low - quality of the powder patterns , more detailed crystallographic data such as anion height and @xmath6fe@xmath6 bond angles are difficult to gather and compare presently .
therefore , more detailed and systematic studies correlating these key parameters should be carried out in future studies .
the consensus so far is that the intercalates should separate the fese layers sufficiently in order to enhance the two - dimensionality of the electronic structure and also charge dope it enough to enhance the coupling between the electron and hole pockets of the fermi surface .
the data from the studies reviewed here suggest that the upper limit for bulk intercalated - fese is 45 k. [ parameters ] from a historical perspective , it is interesting to compare the intercalated iron chalcogenides with that of other 2d hosts such as the transition metal dichalcogenides ( see section [ mch2 ] ) and the na@xmath76coo@xmath12 hydrates .
there seems to be sufficient evidence across these 2d materials that hydrogen bonding has a large role to play in their stability .
the ammonia molecules in both fese and the dichalcogenides such a tas@xmath12 are oriented so that the n h bonds and not the lone pair of electrons on n point towards the chalcogenide anions . in a neutron diffraction study of nh@xmath11 intercalated tas@xmath12 , young et al . found evidence for hydrogen bonding in the form of n
h@xmath108s bonds where the s
h interatomic distances ranged between 2.49 and 2.66 .@xcite similarly , burrard - lucas et al .
found evidence for n
h@xmath108se bridges in the li - nh@xmath11 intercalated fese.@xcite an interesting difference between the iron chalcogenides and the transition metal dichalcogenides is that in the iron chalcogenides the ammonia does not enter the host structure as the conjugate acid ( nh@xmath16 ) , but rather as the conjugate base nh@xmath59 along with neutral ammonia molecules .
favoring amide formation rather than ammonium makes sense in light of the fact that the fese layers are anionic and therefore have electron density at the fermi level to reduce ammonia .
similar to the cobaltates ( see section [ cobaltates ] ) , the iron chalcogenides also have significant aqueous chemistry , and the ability to integrate hydroxides such a lioh .
the weak van der waals forces present in the chalcogenides are presumably replaced by ionic forces from the cationic intercalates .
however , there might be more than these ionic forces participating in interlayer bonding . given the hygroscopic nature of na@xmath76coo@xmath12 and the orientation of the hydroxyl groups in ( li@xmath40fe@xmath17oh)fese , hydrogen bonding should also play a role in the phase stability of these materials . indeed ,
neutron diffraction study by zhou et al .
( li@xmath40fe@xmath17od)fese found crystallographic evidence for o
d@xmath108se bonding where the se
d distance was approximately 3.03 .@xcite the intercalation chemistry of fese also has similarities with the superconducting fullerenes ( see section [ fullerenes ] ) .
the amount of charge doping in both sets of materials seems to be critical to raising the @xmath1 . in the case of intercalated fullerenes , the alkali metals such as potassium electron dope the conduction band and
since this band is narrow , the density of states at the fermi level is high .
similarly , fese requires a certain amount of charge doping at the fermi level to enhance its superconducting properties . unlike the doped fullerenes ,
however , the iron chalcogenides are multi - band superconductors where the fermi surface has both hole and electron pockets instead of just one conduction band .
the nesting of these multiple bands seems to be critical in the case of fese .
even if the maximum @xmath1 has been achieved in fese , there remain multiple future directions for this field .
one , the orientation of the of the organic amines intercalates in fese has not been solved nor the nature of the guest species after intercalation . in other words , are the organic amines neutral or their conjugate bases upon intercalation ?
this question is important to establish the correct electron doping for optimal superconducting properties , which could help answer the question as to the mechanism for such a high @xmath1 .
the coordination of the organic amines , _
e.g. _ pyridine and ethylenediamine , could also be solved in order to determine the importance of hydrogen bonding and the intercalated cations in stabilizing these compounds .
the work of kovnir et al . in the intercalation of structurally related iron selenides with intercalated organic amines demonstrates how this problem could be solved from advanced diffraction techniques.@xcite another set of opportunities in this field is to explore beyond fese .
the intercalation chemistry of mackinawite fes , for example , has not been fully explored . since it has just been found to be superconducting at 5 k
, there exists an opportunity to discover higher @xmath1 sulfides .
finally , other metals besides iron could be investigated such as cobalt , nickel , and copper , which have similar chalcogenide chemistry to iron .
these new intercalated transition metal monochalcogenides could represent a new area of materials chemistry in a similar fashion to the transition metal dichalcogenides .
we thank the national science foundation career , dmr-1455118 , for financial support .
100 url # 1`#1`urlprefixhref # 1#2#2 # 1#1 j. m. carrasco , l. g. franquelo , j. t. bialasiewicz , e. galvn , r. c. p. guisado , m. a. m. prats , j. i. len , n. moreno - alfonso , power - electronic systems for the grid integration of renewable energy sources : a survey , ieee t. ind .
electron . 53 ( 4 ) ( 2006 ) 10021016 .
h. hosono , k. tanabe , e. takayama - muromachi , h. kageyama , s. yamanaka , h. kumakura , m. nohara , h. hiramatsu , s. fujitsu , exploration of new superconductors and functional materials , and fabrication of superconducting tapes and wires of iron pnictides , sci .
tech . adv . mat .
16 ( 3 ) ( 2015 ) 033503 .
s. he , j. he , w. zhang , l. zhao , d. liu , x. liu , d. mou , y .- b .
ou , q .- y .
wang , z. li , l. wang , y. peng , y. liu , c. chen , l. yu , g. liu , x. dong , j. zhang , c. chen , z. xu , x. chen , x. ma , q. xue , x. j. zhou , phase diagram and electronic indication of high - temperature superconductivity at 65 k in single - layer fese films , nat . mater . 12 ( 7 ) ( 2013 ) 605610 .
s. tan , y. zhang , m. xia , z. ye , f. chen , x. xie , r. peng , d. xu , q. fan , h. xu , j. jiang , t. zhang , x. lai , t. xiang , j. hu , b. xie , d. feng , interface - induced superconductivity and strain - dependent spin density waves in fese / srtio@xmath11 thin films , nat . mater .
12 ( 7 ) ( 2013 ) 634640 .
mou , l. zhao , x .-
zhou , structural , magnetic and electronic properties of the iron - chalcogenide a@xmath111fe@xmath44se@xmath112 ( a= k , cs , rb , and tl , etc . ) superconductors , front . phys .
6 ( 2011 ) 410 .
j. l. mertz , z. h. fard , c. d. malliakas , m. j. manos , m. g. kanatzidis , selective removal of cs+ , sr2 + , and ni2 + by k@xmath113mg@xmath111sn@xmath114s@xmath115(x = 0.51 ) ( kms-2 ) relevant to nuclear waste remediation , chem .
25 ( 10 ) ( 2013 ) 21162127 .
a. usuki , m. kawasumi , y. kojima , a. okada , t. kurauchi , o. kamigaito , swelling behavior of montmorillonite cation exchanged for @xmath116-amino acids by @xmath117-caprolactam , j. mater .
res . 8 ( 1993 ) 11741178 .
r. r. chianelli , j. c. scanlon , m. s. whittingham , f. r. gamble , structural studies of the intercalation complexes titanium sulfide - ammonia ( tis@xmath112@xmath10nh@xmath110 ) and tantalum sulfide - ammonia ( tas@xmath112@xmath10nh@xmath110 ) , inorg .
14 ( 7 ) ( 1975 ) 16911696 .
f. r. gamble , b. g. silbernagel , anisotropy of the proton spin - lattice relaxation time in the superconducting intercalation complex tas@xmath112(nh@xmath110 ) : structural and bonding implications , j. chem .
63 ( 6 ) ( 1975 ) 25442552 .
d. w. murphy , f. j. di salvo , g. w. hull , j. v. waszczak , s. f. meyer , g. r. stewart , s. early , j. v. acrivos , t. h. geballe , properties of h@xmath111tas@xmath112 : correlation between the superconducting t@xmath109 and an electronic instability in layer compounds , j. chem .
62 .
p. f. mcmillan , v. cajipe , p. molinie , m. f. quinton , v. gourlaouen , p. colombet , proton dynamics in titanium(iv ) sulfide - ammonia : a new model for ammoniated transition metal dichalcogenides , chem .
3 ( 5 ) ( 1991 ) 796804 . q. huang , m. l. foo , r. a. pascal , j. w. lynn , b. h. toby , t. he , h. w. zandbergen , r. j. cava , coupling between electronic and structural degrees of freedom in the triangular lattice conductor na@xmath111coo@xmath112 , phys .
b 70 ( 2004 ) 184110 .
y. mizuguchi , y. hara , k. deguchi , s. tsuda , t. yamaguchi , k. takeda , h. kotegawa , h. tou , y. takano , anion height dependence of t@xmath109 for the fe - based superconductor , supercond .
( 2010 ) 054013 .
lee , a. iyo , h. eisaki , h. kito , m. t. fernandez - diaz , t. ito , k. kihou , h. matsuhata , m. braden , k. yamada , effect of structural parameters on superconductivity in fluorine - free lnfeaso@xmath70 ( ln = la , nd ) , j. phys .
77 ( 8) ( 2008 ) 083704 .
t. m. mcqueen , q. huang , v. ksenofontov , c. felser , q. xu , h. zandbergen , y. s. hor , j. allred , a. j. williams , d. qu , j. checkelsky , n. p. ong , r. j. cava , extreme sensitivity of superconductivity to stoichiometry in fe@xmath118se , phys .
b 79 ( 2009 ) 014522 .
y. hara , k. takase , a. yamasaki , h. sato , n. miyakawa , n. umeyama , s. i. ikeda , structural and physical properties of fese crystals fabricated by the chemical vapor transport method , physica c 470 , supplement 1 ( 2010 ) s313s314 .
hsu , j .- y .
luo , k .- w .
yeh , t .- k .
chen , t .- w .
huang , p. m. wu , y .- c .
lee , y .- l .
huang , y .- y .
yan , m .- k .
wu , superconductivity in the pbo - type structure @xmath35-fese , p. natl .
usa 105 ( 38 ) ( 2008 ) 1426214264 .
h. kotegawa , s. masaki , y. awai , h. tou , y. mizuguchi , y. takano , evidence for unconventional superconductivity in arsenic - free iron - based superconductor fese : a 77se - nmr study , j. phys .
77 ( 11 ) ( 2008 ) 113703 .
s. margadonna , y. takabayashi , m. t. mcdonald , k. kasperkiewicz , y. mizuguchi , y. takano , a. n. fitch , e. suard , k. prassides , crystal structure of the new fese@xmath40 superconductor , chem .
( 2008 ) 56075609 .
t. m. mcqueen , a. j. williams , p. w. stephens , t. j. , y. zhu , v. ksenofontov , f. casper , c. felser , r. j. cava , tetragonal - to - orthorhombic structural phase transition at 90 k in the superconductor fe@xmath119se , phys .
103 ( 2009 ) 057002 .
a. e. bhmer , t. arai , f. hardy , t. hattori , t. iye , t. wolf , h. v. lhneysen , k. ishida , c. meingast , origin of the tetragonal - to - orthorhombic phase transition in fese : a combined thermodynamic and nmr study of nematicity , phys .
114 ( 11 ) ( 2015 ) 027001 .
m. d. watson , t. k. kim , a. a. haghighirad , n. r. davies , a. mccollam , a. narayanan , s. f. blake , y. l. chen , s. ghannadzadeh , a. j. schofield , m. hoesch , c. meingast , t. wolf , a. i. coldea , emergence of the nematic electronic state in fese , phys .
b 91 ( 2015 ) 155106 .
j. k. glasbrenner , i. i. mazin , h. o. jeschke , p. j. hirschfeld , r. m. fernandes , r. valenti , effect of magnetic frustration on nematicity and superconductivity in iron chalcogenides , nat .
phys . 11 ( 2015 ) 953958 .
a. r. lennie , s. a. t. redfern , p. e. champness , c. p. stoddart , p. f. schofield , d. j. vaughan , transformation of mackinawite to greigite : an in situ x - ray powder diffraction and transmission electron microscope study , am . mineral .
82 ( 302 - 309 ) .
r. gnin , g. bourri , f. trolard , m. abdelmoula , a. jaffrezic , p. refait , v. maitre , b. humbert , , a. herbillon , thermodynamic equilibria in aqueous suspensions of synthetic and natural fe(ii)fe(iii ) green rusts : occurrences of the mineral in hydromorphic soils , envir .
32 ( 8) ( 1998 ) 10581068 .
s. drissi , p. refait , m. abdelmoula , j. gnin , the preparation and thermodynamic properties of fe(ii)/fe(iii ) hydroxide - carbonate ( green rust 1 ) ; pourbaix diagram of iron in carbonate - containing aqueous media , corros .
37 ( 12 ) ( 1995 ) 2025 2041 .
s. j. denholme , s. demura , h. okazaki , h. hara , k. deguchi , m. fujioka , t. ozaki , t. yamaguchi , h. takeya , y. takano , evidence for non - metallic behaviour in tetragonal fes ( mackinawite ) , mater .
147 ( 12 ) ( 2014 ) 5056 .
s. j. denholme , h. okazaki , s. demura , k. deguchi , m. fujioka , t. yamaguchi , h. takeya , m. elmassalami , h. fujiwara , t. wakita , t. yokoya , y. takano , pressure - dependent magnetization and magnetoresistivity studies on tetragonal fes ( mackinawite ) : revealing its intrinsic metallic character , sci . and technol . of adv .
15 ( 5 ) ( 2014 ) 055007 .
m. h. fang , h. m. pham , b. qian , t. j. liu , e. k. vehstedt , y. liu , l. spinu , z. q. mao , superconductivity close to magnetic instability in fe(se@xmath40te@xmath111)@xmath122 , phys . rev .
b 78 ( 2008 ) 224503 .
w. bao , y. qui , q. huang , m. a. green , p. zajdel , m. r. fitzsimmons , m. zhernenkov , s. chang , m. fang , b. qian , e. k. vehstedt , j. yang , h. m. pham , l. spinu , z. q. mao , tunable ( @xmath123@xmath124,@xmath123@xmath124)-type antiferromagnetic order in @xmath35-fe(te , se ) superconductors , phys .
102 ( 2009 ) 247001 .
p. zajdel , p .- y .
hsieh , e. e. rodriguez , n. p. butch , j. d. magill , j. paglione , p. zavalij , m. r. suchomel , m. a. green , phase separation and suppression of the structural and magnetic transitions in superconducting doped iron tellurides , fe@xmath125te@xmath70s@xmath126 , j. am .
( 2010 ) 1300013007 .
t. j. liu , j. hu , b. qian , d. fobes , z. q. mao , w. bao , m. reehuis , s. a. j. kimber , k. prokes , s. matas , d. n. argyriou , a. hiess , a. rotaru , h. pham , l. spinu , y. qiu , v. thampy , a. t. savici , j. a. rodriguez , c. broholm , from ( @xmath124,0 ) magnetic order to superconductivity with ( @xmath124,@xmath124 ) magnetic resonance in fe@xmath127te@xmath40se@xmath111 , nat . mater . 9 ( 9 ) ( 2010 ) 716720 .
e. e. rodriguez , c. stock , p .- y .
hsieh , n. p. butch , j. paglione , m. a. green , chemical control of interstitial iron leading to superconductivity in fe@xmath22te@xmath45se@xmath19 , chem .
sci . 2 ( 9 ) ( 2011 ) 1782 . c. stock , e. e. rodriguez , m. green , p. zavalij , j. rodriguez - rivera , interstitial iron tuning of the spin fluctuations in the nonsuperconducting parent phase fe@xmath22te , phys . rev .
b 84 ( 4 ) ( 2011 ) 045124 .
e. e. rodriguez , c. stock , p. zajdel , k. l. krycka , c. f. majkrzak , p. zavalij , m. a. green , magnetic - crystallographic phase diagram of the superconducting parent compound fe@xmath22te , phys .
b 84 ( 6 ) ( 2011 ) 064403 .
e. e. rodriguez , d. a. sokolov , c. stock , m. a. green , o. sobolev , j. a. rodriguez - rivera , h. cao , a. daoud - aladine , magnetic and structural properties near the lifshitz point in fe@xmath22te , phys .
b 88 ( 16 ) ( 2013 ) 165110 . c. stock , e. e. rodriguez , o. sobolev , j. a. rodriguez - rivera , r. a. ewings , j. w. taylor , a. d. christianson , m. a. green , soft striped magnetic fluctuations competing with superconductivity in fe@xmath22te , phys
b 90 ( 12 ) ( 2014 ) 121113 .
fang , h .-
wang , c .- h .
dong , z .- j .
li , c .- m . feng , j. chen , h. q. yuan , fe - based superconductivity with t@xmath109 = 31 k bordering an antiferromagnetic insulator in ( tl , k ) fe@xmath111se@xmath112 , europhys . lett .
94 ( 2 ) ( 2011 ) 27009 .
r. hu , k. cho , h. kim , h. hodovanets , w. e. straszheim , m. a. tanatar , r. prozorov , s. l. budko , p. c. canfield , anisotropic magnetism , resistivity , london penetration depth and magneto - optical imaging of superconducting k@xmath129fe@xmath130se@xmath112 single crystals , supercond .
24 ( 6 ) ( 2011 ) 065006 .
h. lei , m. abeykoon , e. s. bozin , k. wang , j. b. warren , c. petrovic , phase diagram of k@xmath111fe@xmath44se@xmath131s@xmath132 and the suppression of its superconducting state by an fe@xmath112se / s tetrahedron distortion , phys .
107 ( 13 ) ( 2011 ) 137002 .
a. f. wang , j. j. ying , y. j. yan , r. h. liu , x. g. luo , z. y. li , x. f. wang , m. zhang , g. j. ye , p. cheng , z. j. xiang , x. h. chen , superconductivity at 32 k in single - crystalline rb@xmath111fe@xmath44se@xmath112 , phys .
b 83 ( 2011 ) 060512 .
wang , c .- h .
dong , z .- j .
li , q .- h .
mao , s .- s .
zhu , c .- m .
feng , h. q. yuan , m .- h .
fang , superconductivity at 32 k and anisotropy in tl@xmath133rb@xmath134fe@xmath135se@xmath112 crystals , europhys .
93 ( 4 ) ( 2011 ) 47004 . j. j. ying , x. f. wang , x. g. luo , a. f. wang , m. zhang , y. j. yan , z. j. xiang , r. h. liu , p. cheng , g. j. ye , x. h. chen , superconductivity and magnetic properties of single crystals of k@xmath136fe@xmath137se@xmath112 and cs@xmath138fe@xmath139se@xmath112 , phys .
b 83 ( 2011 ) 212502 .
d. m. wang , j. b. he ,
xia , g. f. chen , effect of varying iron content on the transport properties of the potassium - intercalated iron selenide k@xmath111fe@xmath44se@xmath112 , phys .
b 83 ( 2011 ) 132502 .
l. li , z. r. yang , z. t. zhang , w. tong , c. j. zhang , s. tan , y. h. zhang , coexistence of superconductivity and magnetism in k@xmath111fe@xmath44se@xmath131s@xmath132 ( z = 0,0.4 ) , phys . rev .
b 84 ( 2011 ) 174501 . , luo , x. g. , zhang , m. , wang , a. f. , ying , j. j. , wang , x. f. , yan , y. j. , xiang , z. j. , cheng , p. , ye , g. j. , li , z. y. , chen , x. h. , coexistence of superconductivity and antiferromagnetism in single crystals a@xmath74fe@xmath140se@xmath112 ( a = k , rb , cs , tl / k and tl / rb ) : evidence from magnetization and resistivity , europhys . lett .
94 ( 2 ) ( 2011 ) 27008 .
v. y. pomjakushin , d. v. sheptyakov , e. v. pomjakushina , a. krzton - maziopa , k. conder , d. chernyshov , v. svitlyk , z. shermadini , iron - vacancy superstructure and possible room - temperature antiferromagnetic order in superconducting cs@xmath126fe@xmath141se@xmath112 , phys .
b 83 ( 2011 ) 144410 .
d. p. shoemaker , d. y. chung , h. claus , m. c. francisco , s. avci , a. llobet , m. g. kanatzidis , phase relations in k@xmath111fe@xmath44se@xmath112 and the structure of superconducting k@xmath111fe@xmath112se@xmath112 via high - resolution synchrotron diffraction , phys .
b 86 ( 2012 ) 184511 . x. ding , d. fang , z. wang , h. yang , j. liu , q. deng , g. ma , c. meng , y. hu , h .- h .
wen , influence of microstructure on superconductivity in k@xmath111fe@xmath140se@xmath112 and evidence for a new parent phase k@xmath112fe@xmath142se@xmath143 , nat .
commun . 4 ( 2013 ) 1897 .
p. zavalij , w. bao , x. f. wang , j. j. ying , x. h. chen , d. m. wang , j. b. he , x. q. wang , g. f. chen , p .- y .
hsieh , q. huang , m. a. green , structure of vacancy - ordered single - crystalline superconducting potassium iron selenide , phys .
b 83 ( 2011 ) 132509 .
w. bao , q .- z .
huang , c. gen - fu , m. a. green , w. du - ming , h. jun - bao , q. yi - ming , a novel large moment antiferromagnetic order in k@xmath74fe@xmath144se@xmath112 superconductor , chinese phys .
28 ( 8) ( 2011 ) 086104 .
v. svitlyk , d. chernyshov , e. pomjakushina , a. krzton - maziopa , k. conder , v. pomjakushin , v. dmitriev , temperature and pressure evolution of the crystal structure of a@xmath111(fe@xmath145se)@xmath112 ( a = cs , rb , k ) studied by synchrotron powder diffraction , inorg .
50 ( 21 ) ( 2011 ) 1070310708 . y. zhang , l. x. yang , m. xu , z. r. ye , f. chen , c. he , h. c. xu , j. jiang , b. p. xie , j. j. ying , x. f. wang , x. h. chen , j. p.
hu , m. matsunami , s. kimura , d. l. feng , nodeless superconducting gap in a@xmath111fe@xmath112se@xmath112 ( a = k , cs ) revealed by angle - resolved photoemission spectroscopy , nat .
10 ( 4 ) ( 2011 ) 273277 .
m. burrard - lucas , d. g. free , s. j. sedlmaier , j. d. wright , s. j. cassidy , y. hara , a. j. corkett , t. lancaster , p. j. baker , s. j. blundell , s. j. clarke , enhancement of the superconducting transition temperature of fese by intercalation of a molecular spacer layer , nat .
mater . 12 ( 1 ) ( 2013 ) 1519 .
scheidt , v. hathwar , d. schmitz , a. dunbar , w. scherer , f. mayr , v. tsurkan , j. deisenhofer , a. loidl , superconductivity at t@xmath109 = 44 k in li@xmath111fe@xmath112se@xmath112(nh@xmath110)@xmath126 , eur .
j. b 85 ( 8) .
y. tian - ping , w. gang , j. shi - feng , s. shi - jie , z. han , z. ting - ting , l. xiao - fang , w. wan - yan , c. xiao - long , exploring fese - based superconductors by liquid ammonia method , chinese phys .
b 22 ( 8) ( 2013 ) 087412 .
s. j. sedlmaier , s. j. cassidy , r. g. morris , m. drakopoulos , c. reinhard , s. j. moorhouse , d. ohare , p. manuel , d. khalyavin , s. j. clarke , ammonia - rich high - temperature superconducting intercalates of iron selenide revealed through time - resolved in situ x - ray and neutron diffraction , j. am .
136 ( 2 ) ( 2014 ) 630633 .
s. i. shylin , v. ksenofontov , s. j. sedlmaier , s. j. clarke , s. j. cassidy , g. wortmann , s. a. medvedev , c. felser , intercalation effect on hyperfine parameters of fe in fese superconductor with t@xmath109 = 42 k , europhys .
109 ( 6 ) ( 2015 ) 67004 .
t. p. ying , x. l. chen , g. wang , s. f. jin , t. t. zhou , x. f. lai , h. zhang , w. y. wang , observation of superconductivity at 30@xmath14646k in a@xmath111fe@xmath112se@xmath112 ( a = li , na , ba , sr , ca , yb , and eu ) , sci .
rep . 2 ( 2012 ) 426 .
l. zheng , m. izumi , y. sakai , r. eguchi , h. goto , y. takabayashi , t. kambe , t. onji , s. araki , t. c. kobayashi , j. kim , a. fujiwara , y. kubozono , superconductivity in ( nh@xmath110)@xmath126cs@xmath147fese , phys .
b 88 ( 2013 ) 094521 .
f. r. foronda , s. ghannadzadeh , s. j. sedlmaier , j. d. wright , k. burns , s. j. cassidy , p. a. goddard , t. lancaster , s. j. clarke , s. j. blundell , robustness of superconductivity to structural disorder in sr@xmath19(nh@xmath112)@xmath126(nh@xmath110)@xmath70fe@xmath112se@xmath112 , phys .
b 92 ( 2015 ) 134517 .
l. zheng , x. miao , y. sakai , m. izumi , h. goto , s. nishiyama , e. uesugi , y. kasahara , y. iwasa , y. kubozono , emergence of multiple superconducting phases in ( nh@xmath110)@xmath126m@xmath111fese ( m : na and li ) , sci .
rep . 5 ( 2015 ) 12774 .
m. izumi , l. zheng , y. sakai , h. goto , m. sakata , y. nakamoto , h. l. t. nguyen , t. kagayama , k. shimizu , s. araki , t. c. kobayashi , t. kambe , d. gu , j. guo , j. liu , y. li , l. sun , k. prassides , y. kubozono , emergence of double - dome superconductivity in ammoniated metal - doped fese , sci .
rep . 5 ( 2015 ) 94777 .
k. v. yusenko , j. sottmann , h. emerich , w. a. crichton , l. malavasi , s. margadonna , hyper - expanded interlayer separations in superconducting barium intercalates of fese , chem .
51 ( 2015 ) 71127115 .
d. guterding , h. o. jeschke , p. j. hirschfeld , r. valent , unified picture of the doping dependence of superconducting transition temperatures in alkali metal / ammonia intercalated fese , phys .
b 91 ( 2015 ) 041112 .
y. sakai , l. zheng , m. izumi , k. teranishi , r. eguchi , h. goto , t. onji , s. araki , t. c. kobayashi , y. kubozono , superconducting phases in ( nh@xmath110)@xmath126m@xmath111fese@xmath90te@xmath132 ( m = li , na , and ca ) , phys .
b 89 ( 2014 ) 144509 .
t. hatakeda , t. noji , s. hosono , t. kawamata , m. kato , y. koike , resistive superconducting transition and effects of atmospheric exposure in the intercalation superconductor a@xmath111(c@xmath112h@xmath143n@xmath112)@xmath126fe@xmath131se@xmath112 ( a = li , na ) , j. phys .
568 ( 2 ) ( 2014 ) 022032 .
m. m. hrovat , p. jegli , m. klanj ek , t. hatakeda , t. noji , y. tanabe , t. urata , k. k. huynh , y. koike , k. tanigaki , d. ar on , enhanced superconducting transition temperature in hyper - interlayer - expanded fese despite the suppressed electronic nematic order and spin fluctuations , phys .
b 92 ( 9 ) ( 2015 ) 094513 . a. krzton - maziopa , e. v. pomjakushina , v. y. pomjakushin , f. von rohr , a. schilling , k. conder , synthesis of a new alkali metal organic solvent intercalated iron selenide superconductor with t@xmath109 @xmath149 45 k , j. phys .
condens . mat .
24 ( 38 ) ( 2012 ) 382202 .
t. hatakeda , t. noji , t. kawamata , m. kato , y. koike , new li - ethylenediamine - intercalated superconductor li@xmath111(c@xmath112h@xmath143n@xmath112)@xmath126fe@xmath131se@xmath112 with t@xmath109 = 45 k , j. phys .
82 ( 12 ) ( 2013 ) 123705 .
s. hosono , t. noji , t. hatakeda , t. kawamata , m. kato , y. koike , new intercalation superconductor li@xmath111(c@xmath115h@xmath98n@xmath112)@xmath126fe@xmath150se@xmath112 with a very large interlayer - spacing and t@xmath109 = 38 k , j. phys .
83 ( 11 ) ( 2014 ) 113704 .
t. noji , t. hatakeda , s. hosono , t. kawamata , m. kato , y. koike , synthesis and post - annealing effects of alkaline - metal - ethylenediamine - intercalated superconductors a@xmath111(c@xmath112h@xmath143n@xmath112)@xmath126fe@xmath150se@xmath112 ( a = li , na ) with t@xmath109 = 45 k , physica c 504 ( 2014 ) 811 . c. pak , s. kamali , j. pham , k. lee , j. t. greenfield , k. kovnir , chemical excision of tetrahedral fese@xmath112 chains from the superconductor fese : synthesis , crystal structure , and magnetism of fe@xmath110se@xmath121(en)@xmath112 , j. am .
135 ( 51 ) ( 2013 ) 1911119114 .
x. f. lu , n. z. wang , g. h. zhang , x. g. luo , z. m. ma , b. lei , f. q. huang , x. h. chen , superconductivity in lifeo@xmath112fe@xmath112se@xmath112 with anti - pbo - type spacer layers , phys .
b 89 ( 2014 ) 020507 .
u. pachmayr , f. nitsche , h. luetkens , s. kamusella , f. brckner , r. sarkar , h .- h .
klauss , d. johrendt , coexistence of 3d - ferromagnetism and superconductivity in [ ( li@xmath151fe@xmath111)oh](fe@xmath70li@xmath126)se , angew .
int . edit .
54 ( 1 ) ( 2015 ) 293297 .
x. f. lu , n. z. wang , h. wu , y. p. wu , d. zhao , x. z. zeng , x. g. luo , t. wu , w. bao , g. h. zhang , f. q. huang , q. z. huang , x. h. chen , coexistence of superconductivity and antiferromagnetism in ( li@xmath74fe@xmath73)ohfese , nat .
14 ( 3 ) ( 2015 ) 325329 .
j. w. lynn , x. zhou , c. k. h. borg , s. r. saha , j. paglione , e. e. rodriguez , neutron investigation of the magnetic scattering in an iron - based ferromagnetic superconductor , phys . rev .
b 92 ( 2015 ) 060510 . x. dong , k. jin , d. yuan , h. zhou , j. yuan , y. huang , w. hua , j. sun , p. zheng , w. hu , y. mao , m. ma , g. zhang , f. zhou , z. zhao , ( li@xmath152fe@xmath153)ohfe@xmath154se superconductor : ion - exchange synthesis of large single - crystal and highly two - dimensional electron properties , phys .
b 92 ( 2015 ) 064515 .
h. sun , d. n. woodruff , s. j. cassidy , g. m. allcroft , s. j. sedlmaier , a. l. thompson , p. a. bingham , s. d. forder , s. cartenet , n. mary , s. ramos , f. r. foronda , b. h. williams , x. li , s. j. blundell , s. j. clarke , soft chemical control of superconductivity in lithium iron selenide hydroxides li@xmath40fe@xmath111(oh)fe@xmath70se , inorg .
54 ( 4 ) ( 2015 ) 19581964 .
x. zhou , c. k. h. borg , j. w. lynn , s. r. saha , j. paglione , e. e. rodriguez , the preparation and phase diagrams of ( @xmath155li@xmath40fe@xmath17od)fese and ( li@xmath40fe@xmath17oh)fese , arxiv:1512.03399 .
x. dong , h. zhou , h. yang , j. yuan , k. jin , f. zhou , d. yuan , l. wei , j. li , x. wang , g. zhang , z. zhao , phase diagram of ( li@xmath40fe@xmath111)ohfese : a bridge between iron selenide and arsenide superconductors , j. am .
137 ( 1 ) ( 2015 ) 6669 .
u. pachmayr , d. johrendt , [ ( li@xmath74fe@xmath73)oh]fes and the ferromagnetic superconductors [ ( li@xmath74fe@xmath73)oh]fe(s@xmath40se@xmath111 ) ( 0 @xmath156 x @xmath156 1 ) , chem .
51 ( 2015 ) 46894692 .
x. f. lu , n. z. wang , x. g. luo , g. h. zhang , x. l. gong , f. q. huang , x. h. chen , superconductivity and phase diagram of ( li@xmath74fe@xmath73 ) ohfese@xmath40s@xmath111 , phys .
b 90 ( 2014 ) 214520 .
j. t. greenfield , c. pak , s. kamali , k. lee , k. kovnir , control over connectivity and magnetism of tetrahedral fese@xmath112 chains through coordination fe - amine complexes , chem .
51 ( 2015 ) 53555358 . | the iron chalcogenides fese and fes are superconductors composed of two - dimensional sheets held together by van der waals interactions , which makes them prime candidates for the intercalation of various guest species .
we review the intercalation chemistry of fese and fes superconductors and discuss their synthesis , structure , and physical properties .
before we review the latest work in this area , we provide a brief background on the intercalation chemistry of other inorganic materials that exhibit enhanced superconducting properties upon intercalation , which include the transition metal dichalcogenides , fullerenes , and layered cobalt oxides . from past studies of these intercalated superconductors ,
we discuss the role of the intercalates in terms of charge doping , structural distortions , and fermi surface reconstruction .
we also briefly review the physical and chemical properties of the host materials mackinawite - type fes and @xmath0-fese .
the three types of intercalates for the iron chalcogenides can be placed in three categories : 1 . )
alkali and alkaline earth cations intercalated through the liquid ammonia technique ; 2 . )
cations intercalated with organic amines such as ethylenediamine ; and 3 . )
layered hydroxides intercalated during hydrothermal conditions .
a recurring theme in these studies is the role of the intercalated guest in electron doping the chalcogenide host and in enhancing the two - dimensionality of the electronic structure by spacing the fese layers apart .
we end this review discussing possible new avenues in the intercalation chemistry of transition metal monochalcogenides , and the promise of these materials as a unique set of new inorganic two - dimensional systems .
iron chalcogenides , superconductors , magnetic structure |
the appearance of the effective attractive interactions between electrons in metals @xcite is responsible for the instability of the fermi surface and the resulting phenomenon of superconductivity ; one of the most spectacular quantum effects on a macroscopic scale .
the strong local electron
phonon interaction may induce formation of polaron or bipolaron quasiparticles which are at the heart of bipolaronic theory of superconductivity @xcite .
the so - called negative u centers @xcite are not only a source of superconducting instability but also play an important role in physics and chemistry of materials @xcite ( see , however @xcite ) . the tendency towards bipolaron formation
is enhanced @xcite in the confined structures like quantum dots@xcite .
this assertion has been corroborated by means of quantum monte carlo calculations in the strong coupling regime for various dimensionalities of nano - structures @xcite .
negative charging energy of the small systems like quantum dots is a relatively new concept @xcite .
very small system is characterized by the small capacitance @xmath1 which makes the charging energy @xmath2 large . to overcome it
sufficiently strong polaronic shifts resulting from strong coupling of vibrational and charge degrees of freedom in a molecule are needed .
the charge fluctuations in a device with metallic and superconducting grains may also over - screen the kondo repulsion @xcite in the normal grain .
the purpose of this paper is to extend the recent work @xcite in which charge transport through the molecule @xcite characterized by the negative charging energy has been studied .
modeling the system using anderson like hamiltonian with a negative value of the charging energy @xmath3 and considering weak coupling between the molecule and the electrodes the authors have systematically accounted for the processes contributing to charge transport and calculated its differential conductance .
the study of the negative u molecule has later been extended to the stronger coupling and the regime of the charge kondo effect@xcite .
here we consider the same model and study the charge and heat transport _ via _ negative u molecule in the weak coupling regime when the pair tunneling processes dominate the transport .
the analytical formulas for the thermopower @xmath4 and the thermal conductivity @xmath5 in the linear regime of the low bias voltage @xmath6 and the temperature difference @xmath7 have been found .
@xmath8 denotes the chemical potential and @xmath9 temperature of the left ( right ) external metallic bulk electrode .
@xmath10 denotes the positive electron charge . as in the previous work
we assume that the molecule is characterized by single , doubly degenerated electronic level . in the nonlinear regime with large temperature differences @xmath11 and asymmetric coupling to the external leads the system shows heat rectification property .
the thermal current and conductance in the forward direction ( say , @xmath12 ) differ from those in the reverse direction ( @xmath13 ) .
the low temperature thermal conductance @xmath5 in the linear regime is proportional to @xmath14 , but the heat quantum , defined as @xmath15 takes on a non - universal value .
however , the wiedemann - franz law is obeyed in the limit @xmath16 . at elevated temperatures
we observe marked departures from the fermi liquid behavior with the ratio @xmath17 exceeding the lorenz number @xmath18 .
the organization of the rest of the paper is as follows . in the next section
we briefly recall the model and approach . in section
iii we present the transport coefficients calculated in the linear regime . in section iv the nonlinear transport coefficients are defined and calculated . in section
v we discuss the validity of the wiedemann - franz law and quantization of the thermal conductance in the context of pair tunneling .
we end up with summary and conclusions .
the aim of this section is to recall the main results obtained by koch _
_ @xcite and to establish the notation .
one starts with a general system consisting of external leads and a central molecule coupled by tunneling processes .
the electron subsystem on a molecule interacts with bosonic degrees of freedom ( phonons ) , which are eliminated by performing the firsov - lang canonical transformation .
we assume polaronic shifts large enough to make the effective charging energy @xmath19 negative .
we denote it simply @xmath3 .
the motional narrowing of hopping parameters as well as shifts of the dot energy level are absorbed in the definition of the corresponding parameters in the hamiltonian @xmath20 where @xmath21 describes the interacting electrons on the dot . here
@xmath22 is the number operator , and @xmath23 denotes the creation ( annihilation ) operator of a spin @xmath24 electron on the molecule . the external leads ( left - l and right - r ) are characterized by the electron energy spectrum @xmath25 .
the corresponding term in the hamiltonian is given by @xmath26 , with @xmath27 and @xmath28 denoting the chemical potential of the @xmath29 electrode subject to the bias voltage @xmath30 . in the following
we assume the equilibrium value of the chemical potential @xmath31 and @xmath32 .
the coupling between the molecule and leads is governed by the hamiltonian @xmath33 to correctly account for all low energy processes which contribute to the transport in the limit of negative @xmath3 , it is advisable to eliminate @xmath34 by means of the schrieffer - wolff @xcite transformation .
this is valid in the limits @xmath35 and @xmath36 @xcite .
one gets @xcite the effective low energy hamiltonian @xmath37 in which the direct and exchange @xmath38 interactions between the dot and the leads read @xmath39\,,\nonumber\end{aligned}\ ] ] where @xmath40^{-1}-[\varepsilon-\varepsilon_d - u]^{-1}\ , .
$ ] these terms are most important in the study of transport through the quantum dots with positive charging energy , but they do also play a role in the present case of negative @xmath3 quantum dot @xcite and describe single electron co - tunneling processes .
the pair terms read @xcite @xmath41 for negative values of @xmath3 , the pair tunneling terms play a main role because the state with two electrons on the dot is its lowest energy state and the double occupancy of the dot is favorable .
the two electrons may tunnel onto the dot from a single lead or from two different leads .
the single occupation of the dot is not favored for the low voltage bias @xmath42 and the sequential single particle processes are exponentially suppressed . as a result the single particle events can contribute to the transport through negative u molecule _ via _ higher order processes ( co - tunneling ) .
the approach is based on the rate equations for the occupation probability @xmath43 , where @xmath44 denotes a state of the molecule . since for negative @xmath3
the single occupation of the molecule is never favorable , one finds either one or two electrons on a molecule .
the energies of these two states are equal if @xmath45 . due to the normalization condition
@xmath46 the rate equations reduce to @xmath47 . in a stationary state
this gives @xmath48 , where , in the notation of the paper [ ] , @xmath49 is the total transition rate from the initial state @xmath50 to the final state @xmath51 . to calculate transition rates we use the fermi s golden rule @xcite and consider all processes in which electrons or electron pairs tunnel from the electrodes @xmath29 and @xmath52 to the dot or _ vice versa_. with a constant temperature across the system they have been obtained by koch _
et al . _ and for the process
@xmath53 read @xcite @xmath54 equation ( [ rate1 ] ) describes the transfer rate for tunneling of 2 electrons onto the dot ( spin up electron hops on the dot from the electrode @xmath29 and spin down one from the electrode @xmath52 . in the formula ( [ rate1 ] )
@xmath55 denotes the effective coupling between the dot and the lead @xmath29 , and @xmath56 stands for the density of the states at the fermi level in the electrode @xmath29 .
@xmath57 is the corresponding fermi function .
similar expressions can be derived for all other rates .
the rates of single electron co - tunneling from the left to the right electrode , without changing the occupancy of the dot @xmath58 for the processes @xmath59 and @xmath60 , are obtained @xcite from @xmath38 .
obviously eq .
( [ rate1 ] ) is also valid for system in which the left lead is characterized by temperature @xmath61 different from that of the right lead @xmath62 . in this case , however , no general analytical solution is possible , except in the linear regime .
it is presented in the next section .
first let us consider the transition rate of an electron from a given state in the electrode @xmath29 into the quantum dot as induced by the co - tunneling term in the hamiltonian .
let the initial state @xmath63 of the dot be the vacuum @xmath64 ( no electrons on the dot ) .
the initial state of electrodes represents two fermi surfaces with the electrons occupying the states up to @xmath65 .
we denote the initial state of the whole system by @xmath66 and its energy by @xmath67 .
the process in which an electron with the spin @xmath24 from the state @xmath68 in the electrode @xmath29 is transferred to the state @xmath69 in the electrode @xmath52 results in the final state @xmath70 .
the energy of the final state is @xmath71 . in the single process
the energy transported per unit time from the electrode @xmath29 is @xmath72 .
its contribution to the total energy flux through the left junction equals @xmath73 @xmath74 is the fermi distribution function .
in fact the same factor @xmath72 contributes to the heat flux in an elementary process in which _ e.g. _ a pair is hopping from the states @xmath75 onto the dot , if @xmath76 and @xmath77 . in this process
only the member of the pair with the quantum numbers @xmath78 transports the energy flux through the left junction .
if both electrons stem from the left electrode then their contribution to the total flux reads @xmath79 the conservation of energy expresses the fact that if initially the energy of the system is @xmath80 then final energy is @xmath81 . in calculations of the current flux @xmath82 the extra factor of 2 appears , accounting for a double charge carried in the above process @xcite . the net charge and heat currents ( each of them being a sum of pair and co - tunneling contributions ) in the left junction are given by @xmath83 with @xmath84 , while @xmath85 and completely analogous expressions for @xmath86 and @xmath87 . in the linear regime @xmath88 , @xmath89 and setting @xmath90
we express the transition rates up to the linear order in @xmath91 and @xmath11 , using @xmath92 for the asymmetric couplings @xmath93 the resulting formulae are long and will not be reproduced here . in the special case @xmath94 and for the symmetric distribution of voltages @xmath95 and the temperature difference @xmath96 we obtain @xmath97 the contributions to the charge and heat flux read @xmath98 \label{cond } \\
i_e^t&=&-2\gamma_l\gamma_r{e\over h } m^2(0 ) { \beta x^2\over 4\sinh \beta x } { 1\over t } \\ i_q^v&=&2\gamma_l\gamma_r { e\over h } m^2(0 ) { \beta x^2\over 4\sinh\beta x } \\ k&=&{2\gamma_l\gamma_r\over h } \bigg[\left({f(-x)\over\varepsilon^2_d } + { f(x)\over ( \varepsilon_d + u)^2}\right)i_2 \nonumber \\ & + & m^2(0 ) { \beta x^3\over 8\sinh\beta x}\bigg]{1\over t}\end{aligned}\ ] ] in the above formulae we introduced the notation @xmath99 . @xmath100
measures the distance from the degeneracy point .
it can be tuned by the gate voltage .
note , that the onsager reciprocity relation @xcite @xmath101 is explicitly fulfilled .
it is easy to obtain the phenomenological transport coefficients from those currents .
the linear conductance , @xmath102 , is defined for @xmath103 as @xmath104 and is given by equation ( [ cond ] ) .
it has been discussed earlier @xcite and we shall not discuss it here .
the thermopower @xmath4 is defined as the voltage , which appears across the system subject to the temperature gradient in the absence of current flow @xmath105 s has been studied earlier @xcite and the analytic formula for it reads @xcite @xmath106 thermal conductance @xmath5 is defined as a coefficients between the heat current @xmath107 and the temperature gradient , under the condition of no charge current @xmath108 . with this definition , we find @xmath109\ , .
\end{aligned}\ ] ] in the above formula @xmath110 , denotes the integral @xmath111
outside the linear regime , _
i.e. _ for finite voltage bias @xmath91 and temperature difference between the electrodes , the transfer rates @xmath112 and @xmath113 have to be calculated numerically . in the studies of charge conductance @xcite interesting rectification properties of the device have been predicted . for strongly asymmetric coupling of the dot to the electrodes
e.g. @xmath114 the current @xmath82 is the asymmetric function of @xmath100 in the limit of @xmath115 .
the intriguing question arises if similar rectification properties @xcite could be achieved in the heat flow . to answer this
we have calculated the currents @xmath82 and @xmath107 for general values of @xmath91 and @xmath11 .
formally both charge and heat currents depend on a voltage and a temperature difference @xmath116 as stated earlier the thermal conductance @xmath5 is generally defined by @xmath117 under the condition @xmath118 .
the condition of no current flow defines the thermoelectric power @xmath119 , where @xmath91 is the voltage generated in the system by the temperature gradient .
it provides the selconsistent value of the voltage to be used in equation ( [ heatcur ] ) .
introducing this into the equation for @xmath107 we get @xmath120 the last equality is our definition of the thermal conductance @xmath5 . in this section
we measure all energies in units of @xmath121 . the dependence of @xmath5 on @xmath122 is shown in fig .
( [ fig - kappa ] ) for the temperature @xmath123 and relatively small asymmetry of the junction , parametrised by the value of @xmath124 in the units of @xmath121 , with @xmath125 in the same units . calculating the currents we have assumed symmetric voltage and temperature difference with @xmath126 and @xmath127 .
for positive @xmath11 the heat will normally flow from the left to the right lead .
thermal rectification can be defined @xcite as the dependence of the heat current @xmath128 or the nonlinear thermal conductance @xmath5 defined in eq .
( [ nlthermal - cond ] ) on the sign of temperature difference
. figure ( [ fig - rect ] ) shows the effect of heat rectification in the asymmetric molecular junction with @xmath129 , the temperature @xmath123 and two values of @xmath130 . due to phase space restrictions , which make the transition rates for two electron processes @xmath91 and @xmath11 dependent , the excess heat current @xmath131 depends on @xmath100 and changes sign around @xmath132 for a given set of parameters .
both the heat flux and thermal conductance are the symmetric functions of @xmath100 for small values of asymmetry ( see fig .
( [ fig - kappa]a ) ) .
however , for very strongly asymmetric junctions , the conductance starts to be the asymmetric function of the detuning parameter .
this is shown in the fig .
( [ fig - kappa]b ) .
the rectification coefficient defined as @xmath133 , changes sign as the function of @xmath100 . for the parameters in figure ( [ fig - kappa]a ) it takes the maximal value of about 5% .
it has been predicted @xcite that at low temperature the phononic thermal conductance of one dimensional dielectric wire is universally given by @xmath134 leading to the universal , material and temperature independent ratio @xmath135 , which sets a fundamental quantum limit on heat flow @xcite in an analogy to quantized charge conductance .
the quantized value of the heat conductance is expected @xcite in one - dimensional structures independently of the statistics of heat carriers .
so this is valid for phonons , electrons and also particles with fractional statistics @xcite .
this result has been experimentally @xcite confirmed for phonons , electrons and photons .
it is an easy exercise to show that at @xmath136 the thermal conductance through our negative @xmath3 molecule reduces to @xmath137 thermal conductance , which in the present context consists of only electron contribution is linear in temperature at low temperatures .
the heat quantum , however , is not universal and depends on the coupling amplitudes @xmath138 and the dimensionless distance @xmath100 from the degeneracy point @xmath139 , measured in units of @xmath121 .
it is important to notice that the linear charge conductance at very low temperatures takes on the @xmath14 independent value @xmath140 which ensures the validity of the wiedemann - franz ( wf ) ratio in this limit @xmath141 being one of the signatures of the fermi liquid . for an arbitrary temperature and in the non - linear regime
the above ratio of thermal to charge conductance takes on @xmath14 , @xmath11 and @xmath100 dependent values @xmath142 . the function @xmath143 in units of the lorenz number @xmath144
is shown in figure ( [ fig - wflin ] ) for @xmath145 and @xmath146 . at low @xmath100
it takes values close to 1 , but for larger @xmath100 one observes strong departures from the fermi liquid value @xmath144 , like in the quantum dots with a large coulomb interaction @xcite .
its dependence on the coupling asymmetry is rather weak .
we have also calculated the dimensionless figure of merit which indicates heat to voltage conversion efficiency of the device @xcite .
it is usually defined as @xmath147 . for most devices thermal
conductance contains both electron and phonon contributions . in the present geometry phonons
do not contribute .
the dependence of @xmath148 on @xmath100 for a few values of asymmetry is shown in fig .
( [ fig - zt ] ) .
its value increases with the asymmetry of the coupling and is a non - monotonous function of @xmath100 .
the maximum increases with the anisotropy of coupling as it is evident in the figure .
we have calculated the linear thermoelectric transport coefficients of the single electron molecular transistor in the limit of large electron - phonon coupling leading to a negative effective charging energy . in this limit
the pair tunneling processes contribute mainly to charge and heat transport .
pair dominated transport shows strong departures from the fermi liquid characteristics . we have found strong deviations from the wiedemann - franz law and a non - universal value of thermal conductance quantum .
the thermal quantum _
i.e. _ the ratio @xmath15 evaluated at a low temperature and the lorentz function @xmath143 depend on @xmath100 and asymmetry of the couplings .
the thermoelectric figure of merit @xmath148 characterizes efficiency of the device for thermoelectric applications .
this parameter takes on fairly low , albeit in general @xmath100 and @xmath14 dependent values for the negative @xmath3 quantum dot studied here .
the maximal value of @xmath148 is about 0.25 for the strongly asymmetric junction .
we use the notion quantum dot ( qd ) in a loose sense to describe small central region of the considered structure , independently if it is defined in two dimensional electron gas , consists of a metallic or semiconducting grain or is in the form of a single molecule .
the formula for linear thermopower derived previously @xcite is expressed in terms of the integrals .
unfortunetly , the figures presented in that paper contain numerical error which lead to small departures ( around 2% at the extrema of @xmath4 ) from the exact result .
g. casati , chaos * 15 * , 015120 ( 2005 ) ; b. w. li , l. wang , g. casati , phys .
lett . * 93 * , 184301 ( 2004 ) ; d. segal , a. nitzan , phys . rev
. lett . * 94 * , 034301 ( 2005 ) ; m. terraneo , m. peyrard , g. casati , phys . rev . lett . * 88 * , 094302 ( 2002 ) ; c.w .
chang , d. okawa , a. majumdar , and a. zettl , science * 314 * , 1121 ( 2006 ) ; c. r. otey , w.t .
lau , s. h. fan , phys .
lett . * 104 * 154301 ( 2010 ) ; l. wang and b. li , phys .
world * 21 * , 27 ( 2008 ) ; chen xo , dong b , lei xl , chin .
. lett . * 25 * , 3032 ( 2008 ) .
k. schwab , e. a. henriksen , j. m. worlock , m. l. roukes , nature * 404 * , 974 ( 2000 ) ; o. chiatti , j.t .
nicholls , y.y .
proskuryakov , n. lumpkin , i. farrer , and d. a. ritchie , phys .
lett . * 97 * , 056601 ( 2006 ) ; m. meshke , w. guichard and j. p. pekola , nature * 444 * , 187 ( 2006 ) .
heremans , c.m .
thrush , and d.t .
morelli , phys .
b * 70 * 115334 ( 2004 ) ; j.p .
heremans , acta physica polonica * 108 * , 609 ( 2005 ) ; m. krawiec and k.i .
wysokiski , phys .
b * 73 * , 075307 ( 2006 ) ; m. s. dresselhaus , g. chen , m. y. tang , r. g. yang , h. lee , d. z. wang , z. f. ren , j .-
fleurial , and p. gogna , adv . mat . * 19 * 1043 ( 2007 ) . | charge and heat transport through a single molecule tunnel - coupled to external normal electrodes have been studied .
the molecule with sufficiently strong interaction between electrons and vibrational internal degrees of freedom can be characterized by the negative effective charging energy u@xmath0 . such a molecule has been considered and modeled by the anderson hamiltonian .
the electrical conductance , thermopower and thermal conductance of the system have been calculated as a function of gate voltage in the weak coupling limit within the rate equation approach . in the linear regime
the analytic formulae for the transport coefficients in the pair dominated tunneling are presented .
the effects found in the nonlinear transport include _ inter alia _ the rectification of the heat current .
the sense of forward ( reverse ) direction , however , depends on the tuning parameter and can be controlled by the gate voltage .
we also discuss the quantitation of the thermal conductance and the departures from the wiedemann - franz law . |
quantum mechanics on graphs is recently attracting more and more attention due to its current and prospective use in nanosciences , especially nanoelectronics @xcite .
in particular , since quantum graphs serve as effective models of graph - like structures of submicron sizes , they allow to study single electron devices based on interconnected nanoscale wires .
one of the first such applications emerged in the spectral filtering .
it is known that a line carrying the @xmath0-interaction is usable as a high - pass filter , and similarly , a line carrying the @xmath1-interaction works as a low - pass filter .
more complex spectral filters have been developed as well , for instance a trident filter @xcite , or a y - shaped branching filter , functionning as a high - pass / low - pass junction @xcite . if a quantum device is easily controllable , its usefulness is markedly enhanced . in principle ,
any device based on a quantum graph , including the aforementionned filters , can be controlled by adjusting its internal properties , for example , by varying the parameters of the couplings in its vertices .
however , such a controllability is hard to be implemented in practice since it requires a real - time modification of a nanoscale object .
it is , therefore , highly desirable to develop a device , controllable in a simpler manner . in this paper
we propose a quantum filtering device that can be controlled without affecting its internal structure .
we consider a model in which constant potentials are applied to the lines of a star graph , and the control is achieved through the variation of the potential strengths .
the design of the device essentially relies on a special type of point interaction used in the star graph vertex , namely , on a scale invariant ( flp tsutsui ) coupling .
it should be emphasized that the flp
tsutsui point interaction is very different from usual free coupling and the @xmath0 potential .
its exotic scattering properties together with the external potentials on the lines induce a threshold resonance effect , on which the filtering function is based .
it is known that quantum graph vertices with flp
tsutsui couplings can be approximately constructed from vertices carrying @xmath0-couplings @xcite . in this paper
we improve the existing approximation schemes , and , in addition , demonstrate its convergence in terms of the scattering matrices . since the vertices carrying the @xmath0-couplings can be approximated by regular , smooth potentials @xcite , the realization of the required flp
tsutsui couplings is within reach of the technology in the near future .
therefore , our quantum graph model of a quantum device is of a practical relevance , not just a mathematical abstraction .
the paper is organized as follows : in section [ section : preliminaries ] we prepare the necessary theoretical basis , including the scattering formalism of quantum graphs with external potentials on the lines . in section [ section : device ]
we introduce the concept of a potential - controllable quantum filter .
the first device of this type is designed in section [ section : n=3 ] , namely a band - pass spectral filter with one passband tunable by adjusting the potential on the controlling line . before generalizing the result to more complex devices ,
we devote section [ section : potential ] to a practical question .
although the control potential is assumed to be constant on the whole line , we demonstrate that its small perturbation in a segment close to the vertex , which can be expected in practice , does not distort the filtering function .
the idea of section [ section : n=3 ] is extended to filters with two controlling lines in sections [ section : n=4 ] and [ section : band - pass ] .
the filter proposed in section [ section : n=4 ] has two passbands , and the position of each of them is independently tunable by a control potential .
the filter developed in section [ section : band - pass ] has one passband of adjustable bandwidth : its upper and lower cutoff energies are both tunable by the potentials on the controlling lines . in section [ section : n=2r ]
we introduce a spectral filter with @xmath2 controlling lines , @xmath3 .
the filter has @xmath2 passbands , each of which can be independently adjusted ( as well as disabled ) by the potentials on the controlling lines .
this result is achieved by generalizing the idea of section [ section : n=4 ] .
finally , a filter with multiple independent outputs is studied in section [ section : branching ] .
the outputs work as individually tunable band - pass filters ; the position of the passband at each output is adjustable by tuning a potential on a dedicated controlling line .
the construction of the flp tsutsui couplings is discussed in sections [ section : approximation ] and [ section : s - matrix ] . in section [ section : approximation ] we demonstrate how they can be approximately represented by a small web carrying @xmath0-couplings in the vertices and vector potentials on the lines . in section [ section : s - matrix ] we calculate the on - shell scattering matrix of the approximating graph and prove its convergence to the scattering matrix of the required flp
tsutsui vertex in the small size limit .
the paper is concluded by section [ section : conclusion ] in which we discuss possible extensions and related open problems .
a wave function of a particle confined to a graph having @xmath4 lines @xmath5 of the lengths @xmath6 has @xmath4 components , @xmath7 , and the corresponding hilbert space @xmath8 is given by @xmath9 . let us assume that there are potentials @xmath10 and vector potentials @xmath11 imposed on the graph lines .
the hamiltonian action is given by @xmath12 where @xmath13 is the mass of the particle and @xmath14 is its charge .
note that if there is no vector potential , the hamiltonian takes a simpler form @xmath15 let us consider a graph vertex at which @xmath16 lines are coupled .
we assume that the wave function components at these lines are denoted by @xmath17 and that the coordinates are choosen such that @xmath18 corresponds to the vertex for all @xmath19 .
we introduce the vectors @xmath20 since the hamiltonian is a second - order differential operator , the boundary conditions at the vertex couple @xmath21 and @xmath22 .
the most general form of the boundary conditions is @xmath23 where @xmath24 and @xmath25 are complex @xmath26 matrices . to ensure the self - adjointness of the hamiltonian , which is in physical terms equivalent to conservation of the probability current at the vertex , the matrices @xmath24 and
@xmath25 can not be arbitrary but have to satisfy the requirements @xmath27 where @xmath28 denotes the @xmath29 matrix with @xmath30 forming the first and the second @xmath16 columns , respectively @xcite .
the relation together with the requirements describe all possible vertex boundary conditions giving rise to a self - adjoint hamiltonian .
the requirements can be partly included into the boundary conditions by writing the matrices @xmath30 in a special form .
one of the ways was discovered by harmer @xcite and independently by kostrykin and schrader @xcite who transformed with into `` compact '' boundary conditions @xmath31 where @xmath32 is a unitary @xmath26 matrix .
other way was proposed in @xcite .
it consists in transforming into the block form @xmath33 where @xmath34 , @xmath35 is the identity matrix @xmath26 , @xmath36 is a general complex @xmath37 matrix and @xmath38 is a hermitian matrix of the order @xmath2 .
we call _ @xmath39-form _ of boundary conditions
. it is simple and unique , but on the other hand it requires an appropriately chosen line numbering , see @xcite for details . in this paper
we primarily use the @xmath39-form that best meets our needs .
however , any form of boundary conditions has its pros and cons and is suitable for certain applications .
the @xmath0-coupling ( or `` @xmath0 potential '' ) in a vertex of a quantum graph is characterized by boundary conditions @xmath40 where @xmath41 is a parameter of the coupling .
the @xmath0-coupling is the second most natural singular interaction ( after the free coupling ) by reason of its simple interpretation : it can be understood as a limit case of properly scaled smooth potentials @xcite .
[ delta v ] strictly speaking , the @xmath0 potential of the strength @xmath42 ( @xmath43 ) is characterized by boundary conditions @xmath44 where @xmath13 is the particle mass . however , since the particle in question is almost eclusively electron and , therefore , @xmath45 , the term @xmath46 is commonly written as one single constant @xmath47 , usually called parameter .
if a quantum particle with mechanical energy @xmath48 living on the star graph comes in the vertex from the @xmath49-th line , it is scattered at the vertex into all the lines .
the @xmath50-th component of the final - state wave function is given by @xmath51 where @xmath52 are scattering amplitudes , @xmath53 are angular wavenumbers on the corresponding lines , and coefficients @xmath54 are involved for proper normalization . for any @xmath50 ,
the wavenumber @xmath53 is equal to @xmath55 where @xmath56 is the potential on the @xmath50-th line .
the matrix @xmath57 is the _ scattering matrix _ of the graph . for a normalized wave function coming in from the @xmath50-th line
, @xmath58 is interpreted as the complex amplitude of transmission into the @xmath50-th line ( for @xmath59 ) , whereas @xmath60 represents the complex amplitude of reflection . the matrix @xmath61 depends , in addition to the internal properties of the vertex , on the potentials @xmath62 , and on the particle energy @xmath48 . in order to express this dependence
, we will often use the full symbol @xmath63 in case that @xmath64 for all @xmath19 , we will simplify the notation by omitting @xmath56 , i.e. , @xmath65 in order to derive a formula for @xmath61 , we define matrices @xmath66 and @xmath67 such that @xmath68 , @xmath69 . with regard to , it holds @xmath70 where @xmath71 is the diagonal matrix @xmath72 for @xmath53 given by . any wave function @xmath73 ( the superscript @xmath36 denotes the transposition ) on the graph obeys the boundary conditions determining the vertex . in particular , b. c. must be satisfied by the final - state wave function @xmath74 determined in for all @xmath50 , hence @xmath75 .
when we substitute for @xmath66 and @xmath67 from , we obtain @xmath76 which leads to the sought expression for the scattering matrix @xmath77 : @xmath78 matrix @xmath61 is always unitary , and squared moduli of its elements have the following interpretation : @xmath79 for @xmath80 represents the probability of transmission from the @xmath49-th to the @xmath50-th line , @xmath81 is the probability of reflection on the @xmath49-th line . note that if there are no potentials on the graph lines , i.e. , @xmath64 for all @xmath19 , then @xmath82 for all @xmath50 , and the scattering matrix can be written in the simpler form @xmath83 equation allows to find a relation between the boundary conditions and their form : the boundary conditions are equivalent to @xmath84 it suffices to substitute for @xmath85 from , and then to multiply the equation by the matrix @xmath86 from left .
vertex couplings with @xmath87 independent of @xmath48 are called _ scale invariant _ , or _ flp tsutsui _ , couplings @xcite . in this paper
we use the latter term , since the name `` scale invariant '' sometimes leads to misinterpretations .
it can be easily shown that the scattering matrix @xmath87 corresponding to the boundary conditions is energy - independent if and only if the hermitian matrix @xmath38 in satisfies @xmath88 , i.e. , iff the boundary conditions take the form @xmath89 the scattering matrix corresponding to b.c .
can be explicitely expressed using formula to which we substitute @xmath90 , @xmath91 : @xmath92 this result can be found for example in @xcite .
the @xmath0-interaction can serve for a simple spectral filtering ( fig .
[ d - line ] ) .
this effect can be demonstrated with the help of the scattering matrix .
-interaction on the line as a high - pass spectral filter.,width=170 ] the boundary conditions are equivalent to @xmath93 , hence , using equation , @xmath94 where @xmath95 is the parameter of the @xmath0-interaction .
the transmission amplitude input @xmath96 output ( 1 @xmath96 2 ) is given by @xmath97 , hence we obtain the transmission probability @xmath98 which satisfies @xmath99 to sum up , the system exhibits full reflection for small energies and full transmission for high energies , and , therefore , can be used as a high - pass spectral filter .
the main goal of this paper is to design a quantum filter that is controllable by an external potential , as schematically illustrated in figure [ device ] .
any such device has , therefore , three ( or more ) lines that are used in the following way : * line 1 is _ input_. particles of various energies are coming in the device along this line .
* line 2 is _ output_. particles passed through the device are gathered on this line .
* line 3 is _
controlling line_. we assume that this line is subjected to a constant ( but adjustable ) external potential @xmath100 .
adjustment of the potential directly controls the flow from the input to the output .
the concept can be naturally extended to devices having more than one controlling line , if more device parameters are to be controlled .
similarly , devices with several output lines , or several input lines , can be considered .
we will deal with such cases in sections [ section : n=4][section : branching ] .
2 is being controlled by adjusting a potential @xmath100 on the controlling line 3.,width=170 ] it is desirable that the device is technically simple .
we satisfy this requirement by designing the device modeled by a quantum star graph . in its vertex
, we assume a special , flp tsutsui coupling .
the parameters of the coupling will be specified later .
they determine the transmission characteristics of the device . in order to describe the effect of the adjustable control potentials on the scattering properties of the graph
, we need to generalize the scattering matrix formula .
[ s - matrix ft ] let the vertex coupling in the center of a star graph is given by and there be constant potentials @xmath101 on the graph lines .
then the scattering matrix is given by @xmath102 where @xmath103{1-\frac{v_1}{e}},\ldots,\sqrt[4]{1-\frac{v_r}{e}}\right)\,,\quad q_{(2)}=\operatorname{diag}\left(\sqrt[4]{1-\frac{v_{r+1}}{e}},\ldots,\sqrt[4]{1-\frac{v_n}{e}}\right)\,.\ ] ] the result can be obtained from equation in a similar manner as formula .
the simplest device we are going to construct is based on a star graph with just @xmath104 lines . the flp
tsutsui coupling in its center is described by the boundary condition with two additional assumptions : ( i ) @xmath105 , ( ii ) @xmath36 is real .
the reason for assuming @xmath36 real will become obvious later in section [ section : approximation ] .
hence @xmath106 , @xmath107 , and the boundary conditions read @xmath108 the graph is schematically illustrated in fig .
[ fig : m1 ] .
the roles of individual lines are as stated in section [ section : device ] : 1 = input , 2 = output , 3 = controlling line . on the line 3.,width=170 ] the scattering properties of the graph in question are determined by the scattering matrix @xmath109 , which can be calculated using proposition [ s - matrix ft ] .
if we substitute @xmath106 and @xmath110 , @xmath111 into equation , we easily obtain @xmath112 where @xmath113 ( thus @xmath114{1-u / e}$ ] ) .
a quantum particle with energy @xmath48 coming in the vertex from the input line 1 is scattered at the vertex into all the lines .
the corresponding scattering amplitudes are given by the entries in the first column of .
let us denote for simplicity @xmath115_{j1}$ ] for all @xmath116 .
the transmission amplitude input @xmath96 output equals @xmath117 the reflection amplitude is @xmath118 and the transmission amplitude 1 @xmath96 3 is given by @xmath119{1-\frac{u}{e } } } { 1+a^2 + b^2 \sqrt{1-\frac{u}{e } } } \,\theta(e - u)\,.\ ] ] the heaviside step function @xmath120 is added in equation to make the expression valid for all energies @xmath48 , including @xmath121 .
it represents asymptotically no transmission to the line 3 below the threshold energy @xmath122 .
we are interested above all in the transmission probability from the input line 1 into the output line 2 , which we denote by @xmath123 .
since @xmath124 , we have from @xmath125 \frac{4a^2}{(1+a^2)^2+b^4(u / e-1 ) } & \quad\text{for } e < u . \end{array}\right.\ ] ] we observe that for a given constant potential on the line 3 , @xmath123 as a function of @xmath48 grows in the interval @xmath126 , decreases in the interval @xmath127 , and satisfies @xmath128 therefore , if the parameters @xmath129 are chosen so that @xmath130 , the function @xmath123 has a sharp peak at @xmath131 . furthermore , equation implies that the peak is highest possible ( attaining @xmath132 ) for @xmath133 . to sum up , the device from figure built for @xmath129 satisfying @xmath134 has the following property : the transmission probability input @xmath96 output is high for particles having energies @xmath135 and perfect for @xmath131 , while it is very small for particles with other energies .
it means that such a device works as an adjustable band - pass spectral filter , allowing to control the passband position by the potential put on the controlling line 3 .
the situation is numerically illustrated in figure [ fig : m2 ] .
let be satisfied .
filtering properties of the device , in particular the sharpness of the peak , are determined by the quantity @xmath136 . with regard to the above results , it holds @xmath137 for @xmath138 and @xmath139 for @xmath140 .
the bandwidth @xmath141 of the filter , i.e. , the width of the interval of energies @xmath48 for which @xmath142 , can be calculated from equation .
the bandwidth depends on @xmath100 and @xmath143 , and equals @xmath144 since @xmath145 , it holds @xmath146 . with parameters
@xmath133 , @xmath147 ( i.e. , @xmath148 ) , plotted for the control potential set to @xmath149 .
the transmission probability @xmath123 as a function of @xmath48 is plotted in the left figure .
the right figure shows the reflection probability @xmath150 and the probability of transmission to the controlling line @xmath151.,title="fig:",width=245 ] with parameters @xmath133 , @xmath147 ( i.e. , @xmath148 ) , plotted for the control potential set to @xmath149 .
the transmission probability @xmath123 as a function of @xmath48 is plotted in the left figure .
the right figure shows the reflection probability @xmath150 and the probability of transmission to the controlling line @xmath151.,title="fig:",width=245 ] the resonance at the threshold energy @xmath122 is related to the pole of the scattering matrix which is located on the positive real axis at @xmath152 on the unphysical riemann surface which is connected to the physical riemann surface at @xmath131 .
a _ band - stop _ spectral filter can be constructed in a similar way .
if we begin with @xmath153 and @xmath154 , and choose @xmath155 such that @xmath156 , we obtain @xmath157 hence @xmath158 everywhere except a certain narrow interval around @xmath131 , and @xmath159 .
the filtering property of the device proposed in section [ section : n=3 ] has been derived on the assumption that the control potential @xmath100 is constant on the whole line 3 .
a natural question arises : what happens when the control potential is not constant precisely up to the junction and/or @xmath160 is supported only by a certain segment of the line 3 ? in this section we show that under certain conditions , the filtering property is not distorted . to obtain the result , we prove , for a general potential @xmath160 , that the transmission probability @xmath161 can be expressed in terms of the reflection amplitude calculated for the potential @xmath160 on the real line .
the formula we obtain will allow not only to answer the above question , but also to better understand the function of the filter . in order to find the transmission amplitude @xmath162 for a given potential @xmath160 on the controlling line , we need to apply boundary conditions on the wave function components @xmath163 , @xmath164 , @xmath165 .
the particle motion on the lines 1 and 2 is free , thus the boundary values @xmath166 , @xmath167 at the vertex can be expressed in terms of the scattering amplitudes , cf .
equation , @xmath168 where @xmath169 is the reflection amplitude to the input line 1 , @xmath170 is the transmission amplitude to the output line 2 , and @xmath171 our next aim is to express the boundary values @xmath172 and @xmath173 in a convenient way . for that purpose , let us consider a line with the potential @xmath174 see figure [ fig.potential ] . for @xmath165
being the given wave function component on the controlling line , we define the function @xmath175 where @xmath176 and @xmath177 are determined by the continuity of @xmath178 and @xmath179 at @xmath180 : @xmath181 in the part @xmath182.,width=188 ] the function @xmath178 is obviously a solution of the schrdinger equation for a particle on the line with the potential @xmath183 .
the term @xmath184 corresponding to @xmath185 in definition implies that @xmath178 is the wave function of a particle of energy @xmath48 moving along the line from the part @xmath186 to the part @xmath182 , and @xmath177 represents the reflection amplitude at the point @xmath180 .
[ psi/psi ] from equations it follows @xmath187 .
this formula can be generalized .
it is easy to show that for any @xmath188 , it holds @xmath189 where @xmath190 represents the reflection amplitude corresponding to the shifted and truncated potential @xmath191 , measured at the point @xmath180 .
now we use equations and to subsitute for @xmath192 and @xmath193 , @xmath116 , in the boundary conditions . after an easy manipulation we obtain the system of equations @xmath194 therefore
, the transmission amplitude input @xmath96 output equals @xmath195 if @xmath133 and @xmath136 as introduced in section [ section : n=3 ] , then @xmath196 equation relates the effect of the control potential , encapsulated in the reflection amplitude @xmath177 , and the actual transmission probability @xmath161 of the filter . since @xmath197 by assumption , we observe from equation that the filter is open for the particles that would be reflected by the potential @xmath160 with the amplitude @xmath198 , and closed for particles with @xmath199 .
we can say that the control potential opens the filter by reflecting the particles entering the line 3 back to the vertex with the amplitude @xmath200 .
in particular , if the particle has high enough energy to pass through the potential barrier @xmath160 , we have @xmath201 , hence @xmath202 . in case
that the potential is finitely supported and/or separated from the vertex by a gap , the reflection amplitude @xmath177 is usually oscillating , thus @xmath203 is oscillating as well .
furthermore , if the constant part of the potential is long enough , the oscillations are very rapid and , therefore , imperceptible ( the `` semiclassical '' short wavelength limit ) . in such cases
we mollify the quickly oscillating function @xmath161 by a convolution with the friedrichs mollifier @xmath204 for a certain small @xmath205 , where @xmath206 and @xmath207 is chosen such that @xmath208 .
as we will see in examples below , the obtained mollified function @xmath209 is essentially identical with the transmission probability input @xmath96 output in the ideal case when the control potential is constant on the whole line 3 .
if @xmath211}=u\cdot\chi_{[0,l]}$ ] for @xmath212}$ ] being the characteristic function of the interval @xmath210 $ ] , we have @xmath213 } \mathcal{r}_{[0,l]}=\frac{(1-\xi^2)\sin\xi kl}{(1+\xi^2)\sin\xi kl+2{{\mathrm i}}\xi\cos\xi kl}\,,\ ] ] where @xmath113 .
it can be shown that the transmission probability , given by equation with @xmath177 obeying , grows in the interval of energies @xmath214 $ ] . furthermore
, if the support of @xmath160 is long enough in the sense @xmath215 , then @xmath216 for all particles with energies @xmath217 , and one can prove that * @xmath218})}(e)\approx1 $ ] , * in the region of energies @xmath219 , the transmission probability @xmath220})}(e)$ ] quickly oscillates between @xmath221 and @xmath222 .
the situation is illustrated in figure [ fig .
finite supp ] ( left ) .
the figure shows also the mollified transmission probability @xmath220})}_\mathrm{moll}(e)=(\omega_\epsilon\ast\p^{(u_{[0,l]})})(e)$ ] .
the function @xmath220})}_\mathrm{moll}(e)$ ] is very close to the transmission characteristics obtained in section [ section : n=3 ] for the control potential constant on the whole line 3 .
the condition @xmath223 means that the support of the control potential is substantially longer than the de broglie wavelength of the particle with energy @xmath100 .
such a condition is usually easily satisfied by any macroscopical length @xmath224 . indeed , if we consider an electron and , for instance , @xmath225 ev , then @xmath226 .
, controlled by a potential supported by the finite segment @xmath210 $ ] ( left ) and @xmath227 $ ] ( right ) .
the mollified transmission probabilities are plotted by thick black lines .
the parameters are chosen as @xmath149 , @xmath228 , @xmath229 .
if we assume the particle to be an electron , the energy scale to be given by mev , milielectronvolt , and the length scale to be specified by nm , nanometer , then @xmath230 corresponds to the wavelength @xmath231 . for @xmath232 ,
the mollified transmission probability is essentially identical to the transmission characteristics obtained in the ideal case @xmath233 , cf .
[ fig : m2 ] . , title="fig:",width=245 ] , controlled by a potential supported by the finite segment @xmath210 $ ] ( left ) and @xmath227 $ ] ( right ) .
the mollified transmission probabilities are plotted by thick black lines .
the parameters are chosen as @xmath149 , @xmath228 , @xmath229 .
if we assume the particle to be an electron , the energy scale to be given by mev , milielectronvolt , and the length scale to be specified by nm , nanometer , then @xmath230 corresponds to the wavelength @xmath231 . for @xmath232 ,
the mollified transmission probability is essentially identical to the transmission characteristics obtained in the ideal case @xmath233 , cf .
[ fig : m2 ] . , title="fig:",width=245 ] in the second example we focus on a control potential with the support separated from the vertex by a gap of the length @xmath235 , i.e. , @xmath236}=u\cdot\chi_{[\ell , l]}$ ] ; the corresponding coefficient @xmath237}$ ] equals @xmath238}=\frac{(1-\xi^2)\sin\xi k(l-\ell)}{(1+\xi^2)\sin\xi k(l-\ell)+2{{\mathrm i}}\xi\cos\xi k(l-\ell)}\cdot\e^{2{{\mathrm i}}k\ell}\,.\ ] ] if the lengths @xmath239 satisfy @xmath240 then the values of @xmath237}$ ] are close to the values of @xmath241}$ ] obtained in example 1 ( cf .
indeed , * for @xmath121 or @xmath135 , we have @xmath242 , hence @xmath243 due to the second equation , hence @xmath244 ; * for @xmath245 , it holds @xmath246 , hence @xmath237}\approx0\approx\mathcal{r}_{[0,l]}$ ] . therefore , the transmission characteristics of the filter with @xmath236}=u\cdot\chi_{[\ell , l]}$ ] are similar to those described in example 1 , see figure [ fig
. finite supp ] ( right ) .
namely , the transmission probability @xmath247})}(e)$ ] is quickly oscillating , but the mollified transmission probability @xmath247})}_\mathrm{moll}(e)=(\omega_\epsilon\ast\p^{(u_{[\ell , l]})})(e)$ ] well approximates the ideal characteristics obtained in section [ section : n=3 ] .
the conditions mean that @xmath248 , where @xmath249 is the de broglie wavelength of the particle with energy @xmath100 . in practical implementations , the profile of the control
potential @xmath160 might differ from those studied in examples 1 and 2 above .
in particular , the potential may be constant in the interval @xmath227 $ ] and taper off to zero in the segment @xmath250 $ ] . generally , if the conditions are satisfied , the result is essentially independent of the shape of the fall - off of @xmath160 . to demonstrate this ,
let us distinguish two situations .
* if @xmath121 or @xmath135 , the local wavelength @xmath251 of the particle in every point @xmath252 $ ] satisfies @xmath253 ( the `` long wavelength limit '' ) , hence @xmath254 , @xmath255 .
therefore , with regard to remark [ psi/psi ] , @xmath256}\,,\ ] ] where @xmath257}$ ] is the reflection amplitude corresponding to the potential @xmath258}=u\cdot\chi_{[0,l-\ell]}$ ] . since moreover @xmath259 , it follows from example 1 that the corresponding mollified transmission probability is similar to the transmission characteristics found in section [ section : n=3 ] . * if @xmath245 , the particle is almost fully transmitted through the potential barrier , hence @xmath201 .
therefore , @xmath260 due to equation , which coincides with the result obtained in section [ section : n=3 ] . to sum up
, the filter works well also for a finitely supported potential which is not constant precisely up to the junction , on condition that the support of the potential is long and the segment in which the potential tapers off is short , both with respect to the de broglie wavelength of the particle with energy @xmath100 .
a similar result holds true also for the other filtering devices described in subsequent sections .
in this section we design a band - pass quantum filter with two passbands such that their positions are independently controllable by two external potentials .
the device will be naturally based on a star graph with @xmath261 lines that have the following meaning : 1 = input , 2 = output , 3 and 4 = controlling lines subjected to constant external potentials @xmath262 ; see figure [ fig.4 ] .
the vertex coupling in the graph center is of the flp
tsutsui type in accordance with the concept introduced in section [ section : device ] .
this time we start with an ansatz @xmath153 , i.e. , @xmath263 , and , as in the previous section , assume that @xmath36 is real , thus @xmath264 .
therefore , the vertex is described by the boundary conditions @xmath265 and @xmath266 on the controlling lines 3 and 4 , respectively.,width=170 ] the calculation of the transmission amplitude @xmath267_{21}$ ] can be performed in the same way as in section [ section : n=3 ] using formula ; we find @xmath268 the transmission probability input @xmath96 output is given as @xmath269 .
since the filter we seek shall have two passbands at @xmath131 and @xmath270 , we require zero transmission probability for @xmath140 and high transmission probabilities ( preferably @xmath271 ) at @xmath131 and @xmath270 . the first requirement
( i.e. , @xmath272 ) leads to the condition @xmath273 then we proceed to the requirement @xmath274 , @xmath275 .
let us assume without loss of generality @xmath276 .
since it holds @xmath277 we arrive at the condition @xmath278 conditions and together lead to eight possible expressions for @xmath36 , namely @xmath279 for @xmath280 such that @xmath281 .
we choose the first of them , i.e. , @xmath282 but it is easy to show that all the solutions result in the same transmission probabilities , thus a different choice would not make any difference .
now we use the formula to calculate the corresponding input @xmath96 output transmission amplitude @xmath283 , the reflection amplitude @xmath284 , and the remaining transmission amplitudes @xmath285 , @xmath286 : @xmath287{1-\frac{u}{e}}}{1 + 2a^2\sqrt{1-\frac{u}{e}}}\,\theta(e - u)\,,\\ s_{41}^{(u , v)}(e)&=\frac{2a\sqrt[4]{1-\frac{v}{e}}}{1 + 2a^2\sqrt{1-\frac{v}{e}}}\,\theta(e - v)\ , .
\label{s _ 4,4}\end{aligned}\ ] ] the input @xmath96 output transmission probability is equal to @xmath269 , thus @xmath288 \frac{4a^4}{\left(1 + 4a^4\left(\frac{u}{e}-1\right)\right)\left(1 + 2a^2\sqrt{1-\frac{v}{e}}\right)^2}\,\frac{u - v}{e } & \text { for } v < e < u , \\ [ 1em ] \left(\frac{2a^2\left(\sqrt{1-\frac{u}{e}}-\sqrt{1-\frac{v}{e}}\right)}{\left(1 + 2a^2\sqrt{1-\frac{u}{e}}\right)\left(1 + 2a^2\sqrt{1-\frac{v}{e}}\right)}\right)^2 & \text { for } e > u .
\end{array}\right.\ ] ] if @xmath262 satisfy @xmath289 and @xmath290 , then @xmath291 therefore , the function @xmath292 has two sharp peaks , located at the energies @xmath131 and @xmath270 , see figure [ fig .
p a ] ( left ) .
the situation is different when one of the potentials vanishes .
if @xmath293 and @xmath294 , then @xmath295 therefore , turning one control potential down to zero ( e.g. , @xmath293 ) _ does not _ result in a peak at @xmath296 , but causes the peak corresponding to @xmath266 to vanish , whereas the peak located at @xmath131 remains essentially unaffected ; see figure [ fig .
p a ] for an illustration . to sum up , the device constructed according to the scheme in figure [ fig.4 ] for @xmath36 given by with @xmath297 works as a band - pass spectral filter with @xmath298 passbands .
the positions of the passbands are directly controlled by the external potentials on the controlling lines , and moreover , one band can be suppressed by setting one of the potentials to zero . [ cols="^,^ " , ] [ [ identical - potentials ] ] identical potentials + + + + + + + + + + + + + + + + + + + + if some of the control potentials are of the same nonzero value , or if their values are very close , it has effect on the secondary peaks at that energy .
namely , equation is changed to @xmath299 where @xmath300 can be roughly interpreted as the number of the control potentials that are approximately equal to @xmath301 , i.e. , @xmath302 .
therefore , the height of the secondary peak at @xmath303 indicates the number of control potentials equal to @xmath301 ( or to a certain very close value ) .
let us emphasize that the heights of the principal peaks are essentially unaffected .
[ [ the - effect - of - u_j0 ] ] the effect of @xmath304 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + setting the control potential @xmath305 on the controlling line @xmath306 to @xmath307 essentially results in cancelling the corresponding principal peak .
the limit is changed to @xmath308 and , therefore , the principal peak height is changed from the value given by equation to @xmath309 if all the control potentials are zero , then @xmath310 for all @xmath311 , and the incoming particles are mostly reflected . [
[ another - operation - mode ] ] another operation mode + + + + + + + + + + + + + + + + + + + + + + one may consider the operation mode when the drain line be subjected to a nonzero potential @xmath312 . in such a situation a second principal peak at the energy @xmath313 appears at every output @xmath311 . more precisely speaking , if @xmath314 and satisfies the condition @xmath315 , then @xmath316 whereas the relations are unaffected .
therefore , the device with a nonzero potential at the drain line works as a dual - band branching filter with two principal transmission probability peaks at the outputs : the first peak is at @xmath317 , where @xmath50 denotes the output , the second one is at @xmath313 .
both peaks are essentially of the same height @xmath318 .
the spectral filter designed in section [ section : n=4 ] is obviously a special case of the filter studied in this section , corresponding to @xmath153 .
the function of the controllable filter devices designed in the previous sections is based on the threshold resonance effects in quantum star graphs with `` exotic '' couplings in the vertices , namely those of the flp
tsutsui type .
it should be emphasized that standard vertex couplings ( the free and the @xmath0-coupling ) would not work that way .
it is therefore essential , for the proposed designs to be viable , that the required flp
tsutsui vertices can be physically realized .
this problem has been addressed in @xcite and @xcite , where it was proved that any flp
tsutsui coupling can be approximately constructed by assembling a few short lines in a web with @xmath0-couplings in the vertices and , in some cases , certain vector potentials on the lines .
it essentially solves the problem , because the @xmath0-couplings themselves have a simple physical interpretation and can be well approximated by steep smooth potentials @xcite . in this section
we revisit the results of @xcite and @xcite , proposing a construction which is simpler , and , therefore , easier to be experimentally realized .
the new construction uses less @xmath0-couplings than the scheme from @xcite , and at the same time , in contrast to the scheme from @xcite , it requires no vector potentials in case that the matrix @xmath36 from the boundary conditions is real .
the approximation is constructed in the following way , cf .
figure [ scheme ] : .
certain half lines are connected by short lines ( in this example 23 , 25 , ) , other are without a direct link ( e.g. , 12 , 45 ) .
the bullets represent @xmath0 potentials , the double line symbolizes a vector potential .
the connecting lines are generally of different lengths , and each of them carries either a @xmath0-interactions ( cf .
link 13 ) , or a vector potential ( cf .
14 ) , or none of them ( cf . 15 , 23 , 25 ) , depending on the parameters of the approximated coupling.,width=132 ] * take @xmath16 halflines , each parametrized by @xmath319 , with the endpoints denoted as @xmath320 , @xmath19 .
* for certain pairs @xmath321
( @xmath322 ) , join halfline endpoints @xmath323 by a connecting line of the length @xmath324 , where @xmath235 is a length parameter and @xmath325 is a ( positive ) length coefficient . the pairs @xmath321 such that the vertices @xmath323 are connected , as well as the values of @xmath326 , will be specified later .
if the endpoints @xmath323 are not connected , we may formally set @xmath327 , which represents the `` infinite length '' @xmath328 .
note that @xmath329 for all @xmath330 .
the center of the line connecting @xmath320 and @xmath331 will be denoted by @xmath332 ( therefore , @xmath333 ) .
* put a @xmath0-coupling with the parameter @xmath334 at the vertex @xmath320 for each @xmath335 .
* at certain points @xmath332 ( to be specified later ) place a @xmath0-interaction with a parameter @xmath336 , at the remaining points @xmath332 formally set @xmath337 .
( note that @xmath338 . ) * on certain connecting lines not carrying a @xmath0-interaction ( i.e. , those with @xmath337 ) put a constant vector potential supported by the interval @xmath339 $ ] .
its strength will be denoted by @xmath340 ( if the variable @xmath341 on the line grows from @xmath320 to @xmath331 ) or @xmath342 ( if the variable @xmath341 grows from @xmath331 to @xmath320 ) . both orientation are allowed , and obviously @xmath343 . on the remaining lines
formally set @xmath344 . in the rest of the section
we will specify the dependence of @xmath326 , @xmath334 , @xmath345 and @xmath340 on the matrix @xmath36 and on the length parameter @xmath235 . in order to simplify the notation
, we will write @xmath346 instead of @xmath334 ; the same applies to @xmath347 , @xmath348 .
we will also use symbol @xmath349 standing for the set containing indices of all the lines that are joined to the @xmath50-th one , i.e. @xmath350 . for all @xmath351 , the wave function on the @xmath50-th half line is denoted by the symbol @xmath352 , @xmath353 .
the wave function on the line connecting the points @xmath320 and @xmath331 is denoted by @xmath354 , @xmath355 $ ] . obviously , one has to take the line orientation into account .
we adopt the following convention : if the value @xmath307 corresponds to the endpoint @xmath320 and the value @xmath324 to the endpoint @xmath331 , then the wave function is denoted by @xmath354 . in the case of the opposite direction , the wave function
is denoted by @xmath356 .
therefore , @xmath357 for all @xmath358 $ ] .
now we can proceed to determining @xmath326 , @xmath334 , @xmath345 and @xmath340 . at first , let us express the effect of the @xmath0-couplings involved .
the @xmath0-coupling at the endpoint of the @xmath50-th half line ( @xmath359 ) implies [ ii.serie ] @xmath360 and the @xmath0-interaction in the point @xmath332 means [ ii.serie .. ] @xmath361 where @xmath362 and @xmath363 denote the right - sided limit and the left - sided limit , respectively , of @xmath364 at @xmath341 . further relations which will help us to find the parameters come from taylor expansions .
consider first the case without the vector potentials ( the effect of the potential will be taken into account later ) .
we have @xmath365 @xmath366 where the superscript @xmath307 refers to zero vector potential on the short line connecting @xmath320 and @xmath331 .
if the vector potential on a connecting line is nonzero , the relations and are changed according to the following lemma . [ potencial ]
let @xmath367 , @xmath368 , @xmath369 .
let @xmath370 be the solution of the initial value problem @xmath371 and @xmath372 be the solution of the initial value problem @xmath373 for a certain @xmath374 and a function @xmath375 such that @xmath376 , where @xmath377 denotes the closure of the support of @xmath24 .
then @xmath378 the statement can be proved easily , cf .
@xcite or @xcite .
lemma [ potencial ] implies that the constant vector potential @xmath24 supported by the interval @xmath379 $ ] shifts the phase of the wave function and of its derivative at @xmath380 by @xmath381 .
therefore , equations and are transformed into [ iii.serie ] @xmath382 [ iii.serie .. ] @xmath383 equations , , and connect values of the wave functions and their derivatives at all the vertices @xmath320 , @xmath384 . in the next step
we eliminate the terms with the `` auxiliary '' functions @xmath354 .
to do so , we at first use equations and to obtain @xmath385 then equation gives @xmath386 we substitute for @xmath387 in equation from equation , and find @xmath388 then we apply equation to rewrite the equation above in the form @xmath389 hence we obtain @xmath390 and finally we sum the last equation over @xmath391 , which yields @xmath392 according to equation , the lhs of equation is equal to @xmath393 .
this fact allows us to eliminate the term @xmath394 from equation , which leads to @xmath395 for all @xmath19 .
our objective is to choose @xmath334 , @xmath326 , @xmath345 and @xmath340 in such a way that the system of equations with @xmath19 is equivalent to the boundary conditions in the limit @xmath396 .
it is convenient to adopt the following convention on a shift of the column indices of @xmath36 : [ indt ] the lines of the matrix @xmath36 are indexed from 1 to @xmath2 , the columns are indexed from @xmath397 to @xmath16 .
taking this convention into account , we rewrite the boundary conditions as a system of @xmath16 equations , [ ftft ] @xmath398 the rows of the system are of two types .
we start with the type , corresponding to @xmath399 .
we see immediately that if the parameters @xmath346 , @xmath326 , @xmath347 and @xmath348 are chosen such that @xmath400 then tends to @xmath401 in the limit @xmath396 , and , therefore , equations for @xmath399 are equivalent to equations .
it suffices to find a solution of the system of equations . equation implies that there is no connecting line between @xmath320 and @xmath331 if @xmath402 . if @xmath403 , we use equation ; it is easy to check that it is satisfied by [ spoj ] @xmath404 [ delta xor vector ] for every @xmath399 , the point @xmath320 is connected to a point @xmath331 if and only if @xmath403 and , moreover , @xmath405 .
furthermore , if @xmath320 and @xmath331 are connected , then the @xmath0-interaction in @xmath332 is present iff @xmath406 , and the vector potential on the short line connecting @xmath320 and @xmath331 is present iff @xmath407 .
consequently , none of the short lines connected to @xmath320 , @xmath399 , carries both the @xmath0-coupling and a vector potential .
finally , we use equation together with to find @xmath346 : @xmath408 the case @xmath399 is now solved .
let us proceed to the case @xmath409 .
we substitute for @xmath410 in the left - hand side of equation from equation , and in this way we get @xmath411 then we substitute all the already derived parameters from equations , and into equation . after a manipulation we arrive at @xmath412 similarly as in the case @xmath399 , we require that the system is equivalent to the system in the limit @xmath396 . in this way
we obtain a system of ( sufficient ) conditions on the sought parameters @xmath346 , @xmath413 and @xmath348 ( @xmath409 ) , namely @xmath414 hence we derive explicit expressions for @xmath346 , @xmath326 , @xmath347 and @xmath348 ( @xmath409 ) . from
we have @xmath415 therefore , there is a connecting line between @xmath320 and @xmath331 if and only if @xmath416 .
furthermore , for @xmath416 , we use equation to determine @xmath347 and @xmath417 : [ beta , a j<=m ] * if @xmath418 , then @xmath419 * if @xmath420 , then @xmath421 * if @xmath422 , then @xmath423 note that on none of the short connecting lines there is both the @xmath0-coupling and a vector potential ( similarly as in the case @xmath399 , cf
. remark [ delta xor vector ] ) . finally , we substitute @xmath326 , @xmath347 and @xmath348 found above into equation , and obtain the expression for @xmath346 , @xmath409 : @xmath424 we conclude this section by a practical demonstration .
we use the technique to construct approximations of the flp
tsutsui couplings needed for the filtering devices designed in this paper .
let us consider the coupling used in section [ section : n=3 ] , corresponding to @xmath425 .
since @xmath105 , there is no connecting line between the endpoints @xmath426 and @xmath427 due to equation , cf .
also remark [ delta xor vector ] .
if we substitute @xmath428 , @xmath429 into equation , we obtain @xmath430 , @xmath431 , therefore , the endpoints @xmath432 and @xmath426 are connected by a short line of the length @xmath433 , the points @xmath432 and @xmath427 by a short line of the length @xmath434 , see figure [ fig:6a ] .
since we have assumed @xmath107 , we get @xmath435 from equation ; moreover , the device is contructed for @xmath436 , hence @xmath437 due to equation . therefore
, each short line carries neither a @xmath0-interaction , nor a vector potential .
the parameters of the @xmath0-couplings in the endpoints @xmath426 and @xmath427 follow from equation : @xmath438 , @xmath439 .
finally , the parameter of the @xmath0-coupling in the endpoint @xmath432 is given by equation : @xmath440 .
the quantity @xmath235 appearing in the formulae is the length parameter .
the small size limit @xmath441 , together with the @xmath0 potentials parameters properly scaled according to the formulae above , effectively produces the required flp
tsutsui vertex coupling . and @xmath425 .
for the special choice @xmath133 corresponding to the filter constructed in sect .
[ section : n=3 ] ( cf .
eq . ) , the parameters of the @xmath0 potentials in the points @xmath432 , @xmath426 and @xmath427 are @xmath442 , @xmath443 ( i.e. , there is no @xmath0-interaction in the point @xmath426 ) and @xmath439 .
, width=188 ] the approximation of the coupling corresponding to @xmath444 , used in section [ section : n=4 ] , can be obtained in the same way .
the result is illustrated in figure [ fig:6b ] . and
the parameters of the @xmath0 potentials in the points @xmath432 , @xmath426 , @xmath427 , @xmath445 and @xmath446 are @xmath447 , @xmath448 , @xmath449 , @xmath450 and @xmath451 , respectively.,width=188 ] finally , let us briefly comment on the case when @xmath452 for a unitary @xmath453 matrix @xmath454 ; the corresponding flp tsutsui coupling is essential for constructing the device designed in section [ section : n=2r ] .
it holds : * if @xmath455 , then the endpoints @xmath320 and @xmath331 are not connected due to equation , * if @xmath456 , then the endpoints @xmath320 and @xmath331 are not connected , which follows from the property @xmath457 and equation . consequently , the short lines connect only pairs @xmath320 ,
@xmath331 such that @xmath458 . in other words , they form a bipartite graph in which @xmath459 and @xmath460 are independent sets of vertices , see figure [ fig : bipartite ] .
in addition , if the matrix @xmath454 is chosen as a multiple of an hadamard matrix , then the connecting lines are all of identical length @xmath461 for @xmath235 being the length parameter . and
@xmath452 , @xmath462 .
couplings of this type are used in sect .
[ section : n=2r ] .
the approximating graph in the center has a bipartite structure .
the input and the output lines are on the left hand side , the controlling lines are on the right hand side . some of the connecting lines carry a @xmath0-interaction or a vector potential , depending on the phases of @xmath463.,width=226 ]
now we demonstrate the convergence of the approximating scheme proposed in section [ section : approximation ] .
the proof is performed in terms of the scattering matrices .
we calculate the on - shell scattering matrix of the approximating graph , and show that it tends to the scattering matrix of the approximated flp tsutsui coupling in the limit @xmath396 .
this approach is different from the previous papers on this topic @xcite , where the proof was either omitted , or consisted in performing a full , but very lengthy and cumbersome , demonstration of the norm - resolvent convergence .
in addition , the convergence of the scattering matrices is proved here in a general setting when the lines coupled in the vertex carry potentials @xmath101 , which is another novel point of this paper . in what follows ,
we denote the on - shell scattering matrix of the approximating graph by @xmath464 the matrix @xmath465 will be calculated in the following three steps : 1 .
we derive relations between the on - shell boundary values @xmath466 and derivatives @xmath467 by analyzing the physical properties of the internal web .
we substitute the final - state wave function components @xmath468 , expressed in terms of @xmath465 ( cf .
eq . ) , into the system of equations obtained in the first step .
3 . we solve the system , in this way we find @xmath465 .
_ let @xmath469 and @xmath470 ( @xmath19 , @xmath471 ) be the components of a wave function on the approximated graph , corresponding to a given energy @xmath48 .
they satisfy the boundary conditions in the vertices ; our goal is to derive relations among the boundary values @xmath466 , @xmath467 .
the @xmath0-coupling at the point @xmath320 implies @xmath472 where @xmath473 is the wave function component on the line connecting @xmath320 and @xmath474 , with the convention @xmath180 at @xmath320 and @xmath475 at @xmath474 .
we want to express @xmath476 in terms of @xmath192 and @xmath477 ; the result will be then substituted into equation .
let us assume at first that the short line connecting @xmath320 and @xmath474 carries a @xmath0-interaction with parameter @xmath478 in the center .
the corresponding wave function component is given by @xmath479\ , , \\
[ 1em ] d^+\e^{{{\mathrm i}}kx}+d^-\e^{-{{\mathrm i}}kx } & \text{for all}\quad x\in\left[\frac{d}{2\gamma_{j , h}},\frac{d}{\gamma_{j , h}}\right]\ , , \end{array}\right.\ ] ] where @xmath480 , @xmath481 are certain constants and @xmath482 everywhere in this section , the symbol @xmath391 stands exclusively for the quantity given by equation . from equation , we have [ elimcd ] @xmath483 furthermore , the @xmath0-interaction in the center of the line means @xmath484 hence @xmath485 by solving the system , we find @xmath480 and @xmath481 in terms of @xmath192 and @xmath477
. then we substitute @xmath480 into the equation @xmath486 which follows from , and in this way we find @xmath487 recall that the formula has been derived on the assumption that the short line connecting @xmath320 and @xmath474 carries a @xmath0-interaction with the parameter @xmath478 .
the solution for the case when the line carries a vector potential @xmath488 can be obtained using lemma [ potencial ] .
both solutions can be merged into the general formula @xmath489 which is applicable to any of the three possible situations ( a @xmath0-interaction , a vector potential , none of them ) , it suffices to set @xmath490 or @xmath491 if necessary .
now we sum equations over @xmath492 and apply , in this way we get @xmath16 equations relating @xmath192 and @xmath193 : @xmath493 _ step 2 . _
if a particle with energy @xmath48 comes from the @xmath49-th line into the internal approximating web , the final - state wave function naturally obeys relations .
let @xmath494 denote the components of the final - state wave function on the half lines and @xmath494 , @xmath495 be the boundary values and derivatives .
we express @xmath496 and @xmath497 in terms of the entries of @xmath465 using the definition of the scattering matrix : @xmath498_{\ell\ell}\right ) & \text{for $ j=\ell$ } , \\
\frac{1}{\sqrt{k_j}}[\s_d]_{j \ell } & \text{for $ j\neq \ell$ } , \end{array}\right . \label{rt } \\[1em ] ( \psi^{(\ell)}_j)'(0)=\left\{\begin{array}{cl } { { \mathrm i}}\sqrt{k_\ell}\left(-1+[\s_d]_{\ell\ell}\right ) & \text{for $ j=\ell$ } , \\ { { \mathrm i}}\sqrt{k_j}[\s_d]_{j \ell } & \text{for $ j\neq \ell$ } , \end{array}\right .
\label{rt ' } \end{gathered}\ ] ] where @xmath499 for all @xmath19 .
if we multiply equation by @xmath235 , then for all @xmath19 replace @xmath192 and @xmath193 by @xmath500 and @xmath495 , respectively , and finally , substitute for @xmath500 and @xmath495 from equations and , we obtain a system of @xmath501 equations for @xmath502_{j\ell}$ ] , @xmath503 .
they take different forms for @xmath504 and @xmath505 : [ z jl ] @xmath506case @xmath59 @xmath507_{j \ell}+\\ + kd\sum_{h\in n_j}\frac{\e^{-{{\mathrm i}}\frac{d}{2\gamma_{j , h}}a_{(j , h)}}}{\sin k\frac{d}{\gamma_{j , h}}+\frac{\beta_{j , h}}{k}\sin^2 k\frac{d}{2\gamma_{j , h}}}\frac{1}{\sqrt{k_h}}[\s_d]_{h\ell } \\ = -kd\frac{\e^{-{{\mathrm i}}\frac{d}{2\gamma_{j,\ell}}a_{(j,\ell)}}}{\sin k\frac{d}{\gamma_{j,\ell}}+\frac{\beta_{j,\ell}}{k}\sin^2 k\frac{d}{2\gamma_{j,\ell}}}\frac{1}{\sqrt{k_\ell}}\cdot\chi_{n_j}(\ell)\,,\end{gathered}\ ] ] where @xmath508 for @xmath471 and @xmath509 otherwise .
@xmath506case @xmath505 @xmath510_{\ell\ell}+\\ + kd\sum_{h\in n_\ell}\frac{\e^{-{{\mathrm i}}\frac{d}{2\gamma_{\ell , h}}a_{(\ell , h)}}}{\sin k\frac{d}{\gamma_{\ell , h}}+\frac{\beta_{\ell , h}}{k}\sin^2 k\frac{d}{2\gamma_{\ell , h}}}\frac{1}{\sqrt{k_h}}[\s_d]_{h\ell }
\\ = \left(kd\sum_{h\in n_\ell}\frac{\cos k\frac{d}{\gamma_{\ell , h}}+\frac{\beta_{\ell , h}}{2k}\sin k\frac{d}{\gamma_{\ell , h}}}{\sin k\frac{d}{\gamma_{\ell , h}}+\frac{\beta_{\ell , h}}{k}\sin^2 k\frac{d}{2\gamma_{\ell , h}}}+\alpha_\ell d+{{\mathrm i}}k_\ell d\right)\frac{1}{\sqrt{k_\ell}}\end{gathered}\ ] ] _ step 3 .
_ the sought scattering matrix @xmath465 can be obtained by solving the system of equations .
we observe that the system can be rewritten in the matrix form @xmath511 where @xmath512 is the matrix introduced by equation and @xmath513 therefore , @xmath514 note that the matrices @xmath515 and @xmath71 occuring in equation have very different meanings .
the matrix @xmath515 represents primarily the _ internal _ properties of the approximating graph , and depends on the approximated boundary conditions , on @xmath235 and on @xmath48 .
the matrix @xmath71 represents only the _ external _ setting , and depends on @xmath101 and on @xmath48 .
it suffices to determine @xmath515 .
this will be done by substituting the values of the coupling parameters @xmath346 , @xmath347 , the length factors @xmath326 and the vector potential strengths @xmath348 into equation .
all these parameters are given by equations , , , , .
we shall keep in mind that the expressions are different when @xmath330 belong to @xmath516 and when @xmath330 belong to @xmath517 , and moreover , @xmath518 is defined in different ways for @xmath504 and for @xmath505 . therefore , the formula for @xmath519 is divided into several expressions : [ z form ] @xmath506case @xmath456 and @xmath59 @xmath520 @xmath506case @xmath521 @xmath522 @xmath506case @xmath523 @xmath524 @xmath506case @xmath458 @xmath525 @xmath506case @xmath526 and @xmath59 @xmath527 @xmath506case @xmath528 @xmath529 equations above fully determine the matrix @xmath515 .
therefore , for a given @xmath36 , @xmath101 , @xmath48 and @xmath235 , we can easily find @xmath465 using the formula together with equations , and .
if @xmath396 , i.e. , @xmath530 , we can expand each of the expressions .
the calculation leads to @xmath531 when the matrix is substituted into equation , one obtains the following approximative formula for @xmath465 : @xmath532 where @xmath533 and @xmath534 , @xmath535 are the diagonal matrices introduced in proposition [ s - matrix ft ] . in equation , the scattering matrix of the approximating graph is found up to a small error .
the error tends to zero for @xmath396 , i.e. , @xmath536 now we compare the limit with the scattering matrix of the approximated star graph . since the vertex of the approximated graph carries a flp
tsutsui coupling , described by the boundary conditions , its scattering matrix @xmath77 is given by formula .
comparing equations and , we find @xmath537 to sum up , the scattering matrix of the approximating system converges to the scattering matrix of the requested flp
tsutsui vertex coupling in the limit @xmath396 .
the resonance phenomenon found in this paper allows to control the transmission of a quantum particle along a line by external potentials put on other lines connected to the channel at a vertex .
we have demonstrated that for certain well - chosen flp
tsutsui vertices , the resonance is strong enough to enable a design of quantum spectral filters with fine characteristics .
the scattering characteristics are very different depending on the matrix @xmath36 determining the flp tsutsui coupling in the graph vertex .
therefore , it is likely that many more controllable quantum devices based on this concept can be obtained for other choices of @xmath36 .
some of the designed filters have rather small transmission probabilities inside the passbands
. it would be useful to find how to increase the probabilities , and in this way to improve the filter characteristics .
macroscopic controllability can be achieved also by an application of a magnetic field , for instance , by tuning a magnetic flux through rings carried by the controlling lines , or through ring - shaped controlling lines .
it turns out that the use of suitably chosen flp
tsutsui couplings together with the application of magnetic field lead to transmission characteristics with adjustable high peaks .
we plan to address this problem in a subsequent article .
it seems that the idea can be extended to the control of the spin flow , because the mathematical description of a spin one - half particle propagating in a quantum star graph with @xmath16 lines is equivalent to the description of a spinless particle in a quantum star graph with @xmath538 lines .
exploring the spin - filtering mechanism could be of interest .
finally , there are known parallels between the particle propagation in quantum graphs and the wave propagation in waveguides .
consequently , we are convinced that the results of this paper can be adapted to the control of a transmission of macroscopic electromagnetic microwaves in a thin waveguide .
we thank prof . atushi tanaka for helpful suggestions .
this research was supported by the japan ministry of education , culture , sports , science and technology under the grant number 21540402 .
p. exner , j.p .
keating , p. kuchment , t. sunada , a. teplyaev , eds . : _
analysis on graphs and applications _ ,
ams `` proc . of symposia in pure math . ''
, vol . 77 , providence , r.i . , 2008 , _ and references therein_. m. lawniczak , s. bauch , o. hul and l. sirko : experimental investigation of the enhancement factor for microwave irregular networks with preserved and broken time reversal symmetry in the presence of absorption .
rev . _ * e 81 * , 046204 ( 5pp ) ( 2010 ) .
t. shigehara , h. mizoguchi , t. mishima , t. cheon : realization of a four parameter family of generalized one - dimensional contact interactions by three nearby delta potentials with renormalized strengths , _ ieice trans .
comm . comp .
_ * e82-a * ( 1999 ) , 17081713 . | we study the scattering in a quantum star graph with a flp
tsutsui coupling in its vertex and with external potentials on the lines .
we find certain special couplings for which the probability of the transmission between two given lines of the graph is strongly influenced by the potential applied on another line . on the basis of this phenomenon
we design a tunable quantum band - pass spectral filter .
the transmission from the input to the output line is governed by a potential added on the controlling line .
the strength of the potential directly determines the passband position , which allows to control the filter in a macroscopic manner .
generalization of this concept to quantum devices with multiple controlling lines proves possible .
it enables the construction of spectral filters with more controllable parameters or with more operation modes .
in particular , we design a band - pass filter with independently adjustable multiple passbands .
we also address the problem of the physical realization of flp tsutsui couplings and demonstrate that the couplings needed for the construction of the proposed quantum devices can be approximated by simple graphs carrying only @xmath0 potentials .
quantum graphs , threshold resonance , singular vertex coupling , quantum control 03.65.-w , 03.65.nk , 73.63.nm |
in the last few years there has been a growing interest in quantum information theory , in particular in uses of entanglement like teleportation @xcite , quantum cryptographic key distribution @xcite and dense coding @xcite .
in addition to extensive theoretical work , some tasks involving manipulation of entanglement have been successfully demonstrated in practice , including teleportation @xcite and quantum key distribution @xcite . in this paper
i discuss the feasibility of experimentally implementing another application of quantum entanglement : quantum communication complexity protocols .
the scenario of communication complexity ( cc ) was introduced by yao @xcite , who investigated the following problem involving two separated parties ( alice and bob ) .
alice receives a @xmath0-bit string @xmath1 and bob another @xmath0-bit string @xmath2 , and the goal is for one of them ( say bob ) to compute a certain function @xmath3 with the least amount of communication between them .
of course they can always succeed by having alice send her whole @xmath0-bit string to bob , who then computes the function , but the idea here is to find clever ways of calculating @xmath4 with less than @xmath0 bits of communication .
this problem is relevant in many contexts : in electronic circuit design , for example , one wants to minimize energy use by decreasing the amount of electric signals required between the different components during a distributed computation .
various authors have established @xcite the somewhat surprising result that entanglement can be used to reduce the amount of communication necessary to calculate many functions of distributed inputs .
this can be done by previously sharing entangled states between the parties , and then allowing them to do measurements on these states as part of their cc protocols .
quantum cc tries to quantify the gain obtained by using entanglement for different classes of multi - party distributed function calculations . in this paper
i present a simple two - party cc task and discuss some necessary conditions to demonstrate the quantum protocol in the laboratory .
i start by describing the task itself in section [ sec task ] and its optimal classical protocol ( section [ sec class prot ] ) , followed by a more efficient quantum protocol ( section [ sec quant prot ] ) . in section [ sec bell ]
i show that the efficiency of the quantum protocol relies on its equivalence to measurements that maximally violate the clauser - horne - shimony - holt ( chsh ) quantum non - locality inequality . in section [ sec experim
] i discuss some issues concerning the feasibility of implementing this protocol experimentally , and argue that ion trap technology is presently very close to achieving the necessary thresholds . in section [ sec three party ] i analyse a three - party cc task discussed in @xcite and show that in principle it can be demonstrated experimentally with lower detector efficiencies and higher background count rates than the two - party quantum cc protocol previously discussed . finally in section [ sec conclu ] i round up with some concluding remarks .
in @xcite the authors presented a two - party cc task on which i will now elaborate , introducing a slight change in order to obtain a larger gap in efficiency between the quantum and the classical protocols .
this larger difference will be important when we discuss the experimental implementation in section [ sec experim ] , as it allows for lower detector efficiencies .
the modified task can be simply stated as follows .
the two parties alice and bob are each given a number between @xmath5 and @xmath6 , where @xmath7 is an even integer and @xmath8 .
we can think of alice s number @xmath1 and bob s number @xmath2 as being integers modulo @xmath9 lying uniformly distributed on a circle .
the numbers are chosen randomly with equal probabilities , but obeying certain constraints : either @xmath10 in which case @xmath1 and @xmath2 are said to be ` neighbours ' , or @xmath11 in which case @xmath1 and @xmath2 are ` anti - neighbours ' .
the reason for the terms ` neighbours ' and ` anti - neighbours ' is obvious when one looks at the two possibilities in the integers modulo @xmath9 circle : in the ` neighbours ' case @xmath1 and @xmath2 are adjacent points ; in the ` anti - neighbours ' case @xmath2 is adjacent to the farthest point from @xmath1 , which is @xmath12 . after being assigned their numbers ,
alice is then allowed to communicate a single bit of information to bob , who then has to decide whether @xmath1 and @xmath2 are neighbours or anti - neighbours , achieving as low an error rate as possible over many runs of the task .
this protocol represents only a slight departure from that presented in @xcite .
the difference here is that @xmath1 and @xmath2 are not allowed to coincide or to differ by @xmath13 , as was the case in @xcite . as we will see in section
[ sec quant prot ] , this eliminates two possibilities for which the quantum protocol does not offer advantage over the best classical one , making the quantum protocol more efficient in relation to the optimal classical counterpart . in this section i describe the best possible deterministic classical protocol , and argue that it is optimal . after receiving her number @xmath1
, alice decides about which bit value to send to bob by looking up the value of some function @xmath14 , as @xmath1 is only data she has access to .
the function @xmath15 is also known to bob , who upon receiving the bit from alice learns that @xmath1 lies in one of two disjunct sets , and acts in an optimal way to try to decide whether @xmath1 and @xmath2 are neighbours or antineighbours .
we may think of the function @xmath16 as a colouring of the @xmath9 points in the circle in either black ( corresponding to @xmath17 ) or white ( @xmath18 ) .
what alice does is to send @xmath1 s colour to bob for him to decide which case they have .
suppose , for example , that alice sends the bit @xmath19 to bob , corresponding to a black @xmath1 . if bob s @xmath2 has two black neighbours and two white anti - neighbours , then bob can be absolutely sure that @xmath1 and @xmath2 are neighbours .
of course the situation will not always be so clear - cut , forcing bob to make informed or random guesses sometimes .
one way to find the optimal deterministic classical protocol is to consider all possible @xmath16 ( = colourings ) and evaluate the probability of success for bob , using the fact that the distributions for @xmath1 and @xmath2 are uniform .
it is a reasonable conjecture that one optimal colouring consists of one half of the circle coloured white and the other half black , which yields a success probability of @xmath20 for @xmath21 , i analysed the 16 non - equivalent possible colourings and indeed found that the optimal deterministic protocol is of the form above , having a success probability of@xmath22 even though we considered only deterministic protocols , in @xcite the authors showed that probabilistic protocols need not be considered for this kind of problem .
this is true because any probabilistic protocol can be reduced to a deterministic one by having alice and bob share random numbers before they are assigned @xmath1 and @xmath2 . as a result , for @xmath21
the best possible classical protocol is the one described above , achieving a success probability of @xmath23 . here
i describe a quantum protocol which accomplishes the task with a larger success probability than the optimal classical protocol described above .
the motivation for this protocol comes from the visualization of @xmath1 and @xmath2 on the @xmath24 circle , where they lie separated by an angle which is either @xmath25 or @xmath26 .
it is a well known fact that local rotation followed by measurements on maximally entangled states yield strong correlations which are not allowed classically @xcite , and which will be used in the protocol .
before they are given their numbers , alice and bob meet and share a singlet state @xmath27 for simplicity of description , we will assume that @xmath28 was shared in the form of an entangled pair of spin-@xmath29 particles . upon receiving her number , alice measures the spin on her particle along the axis defined by the angle @xmath30 to the @xmath31 axis on the @xmath32 plane , and sends the result ( @xmath33 ) to bob .
he measures the spin on his particle along the axis defined by the angle @xmath34 to the @xmath31 axis on the @xmath32 plane , also obtaining a result @xmath33 .
it does not matter whether he does this before or after receiving alice s message .
bob s decision is made in a simple way : if the two measurement results are the same he guesses that @xmath1 and @xmath2 are neighbours , otherwise that they are anti - neighbours .
let us now evaluate the probability of error if bob and alice adopt this protocol . for the singlet state one can easily show that@xmath35 where @xmath36same@xmath37 is the probability that their outcomes are
the same and @xmath36opposite@xmath37 is the probability that their outcomes are opposite . in the ` neighbours ' case
we have @xmath38 and hence @xmath39 in the ` anti - neighbours ' case we have either @xmath40 or @xmath41 , from which we find @xmath42 in any case , we see that this protocol is successful with a probability @xmath43 which is always larger than the conjectured classical success probability given by eq .
( [ pc n ] ) , and definitely larger than the best classical protocol for @xmath21 , whose success probability is limited to @xmath23 [ compare with @xmath44 . unlike the classical protocol
, the quantum version has a probability of success which is independent of the probability distribution for @xmath1 and @xmath2 .
this , however , should not be considered a substantial advantage for the quantum protocol , as a randomized classical protocol can be used to obtain @xmath45 even for non - uniform distributions of @xmath1 and @xmath2 .
this can be done by having alice and bob choose different optimal colourings for each run , according to previously shared random numbers . in the protocol described above the singlet state ( [ singlet ] ) is measured along directions that differ by either @xmath25 or @xmath26 .
this is in contrast with the quantum protocol for the task presented in @xcite , whose measurements include those but also measuments along the same axis . since the quantum and classical correlations are equal ( and maximal ) for measurements along the same axis , we obtain a larger gap between the classical and quantum efficiencies by eliminating this possibility , which we did by redefining the task suitably . the quantum protocol described above is clearly inspired by the chsh inequality presented in @xcite , which establishes bounds for classical correlations between simultaneous measurements on maximally entangled states .
in fact , we will see that there is a simple interpretation of the quantum protocol in terms of bell - type quantum non - locality tests . in @xcite an inequality
was derived for classical correlations between measurements with two possible outcomes ( denoted by @xmath46 or @xmath47 ) , arising from experiments with @xmath48 different setups at alice s laboratory and @xmath48 at bob s .
we write @xmath49 to denote the probability of obtaining two equal outcomes ( @xmath50 or @xmath51 ) with setups @xmath52 at alice and @xmath53 at bob , and similarly @xmath54 for different outcomes ( @xmath55 or @xmath56 ) .
for @xmath21 we have the following inequality @xcite : @xmath57 which must be obeyed by _ any _ local theory .
this inequality is saturated by a simple ` classical spin ' local model described in @xcite .
it is easy to see that quantum mechanics violates inequality ( [ hardy ineq . ] ) for measurements on the singlet state ( [ singlet ] ) . choosing the angles of measurement
to be @xmath58 we obtain maximal violation of inequality ( [ hardy ineq . ] ) : @xmath59 we see that the quantum mechanics predictions violate the inequality , and therefore are impossible to be accounted for by any local theory .
the same setup also maximally violates the chsh inequality , as discussed in @xcite ; in fact hardy has shown @xcite that the chsh inequality can be obtained from ( [ hardy ineq . ] ) .
it is interesting to note that there is an exact correspondence between the measurements that maximally violate inequality ( [ hardy ineq . ] ) and the quantum protocol for @xmath21 described above . in both cases
the two parties do measurements on a singlet state along co - planar directions separated by two possible angles : @xmath60 or @xmath61 radians . in the quantum protocol
we want to maximize @xmath62 the inequality test described above corresponds to a similar situation .
quantum mechanics predicts @xmath63 @xmath64neighbours@xmath65 and @xmath66anti - neighbours@xmath65 , and we maximize @xmath67neighbours@xmath68 @xmath64anti - neighbours@xmath65 in order to violate the inequality by the greatest amount@xmath69 as @xmath70neighbours@xmath71anti - neighbours@xmath65 , the maximization of the violation of inequality ( [ ineq viol ] ) also maximizes the probability of success @xmath72 .
curiously , the ` classical spin ' model used to saturate the inequality turns out to be equivalent to the best classical protocol for the task , that is , it can be used as an alternative to the optimal classical protocol described in section [ sec class prot ] .
the key difference between the measurements used in the quantum protocol and the chsh inequality test is the number of possible setups at alice and bob .
for the chsh inequality violation we consider two setups at alice and two at bob , whereas in the quantum protocol alice and bob need to use one of four possible setups .
the ` promise ' that @xmath1 and @xmath2 are either neighbours or antineighbours guarantees that for each run the measurements at alice and bob will be done along angles that differ either by @xmath60 or @xmath61 radians , like the chsh test .
but unlike it , the quantum protocol relies on alice using one of four possible setups ; she can not communicate this to bob with a single bit .
instead , she communicates the result of the measurement to bob , who can then use the stronger - than - classical correlations between the measurements to obtain success with a high probability .
the quantum cc protocol of section [ sec quant prot ] with @xmath21 is arguably the simplest to implement in the laboratory .
there is a clear - cut criterion for experimental success : if after a large number of runs the success rate is found to be larger than @xmath73 then entanglement - enhanced cc will have been demonstrated . in @xcite
a similar cc task was presented which has the same classical and quantum error rates as the one here , and which therefore needs an equivalent detector efficiency to be implemented ( as we will see below ) .
the advantages of the present task is that it is simpler to describe and easier to interpret in terms of bell - type inequalities , as discussed in the previous section .
it is important to point out that in order to experimentally implement this quantum cc protocol it is _ not _ necessary to have space - like separated measurements at alice and bob . unlike bell inequality tests , here the two parties are _ required _ to communicate before finishing the task .
all that is needed for a successful experiment is to implement the quantum protocol in such a way as to obtain a success rate which is higher than that of the highest classical protocol .
once this is done , one might correctly argue that , from a fundamental point of view , local hidden - variable theories can not be ruled out as the responsible for the better performance of the quantum protocol . from a practical point of view , however , after a large enough number of runs this would be an undeniable demonstration of entanglement - assisted communication .
the question of whether local hidden variable theory or quantum mechanics is the correct theory to explain the enhancement becomes immaterial if we are only concerned about establishing an experimental evidence for the effect . given the equivalence of the protocol with a chsh inequality test ,
the experimental problems to be taken into account are those of a chsh test , except that we do not need to close the locality loophole . with this in mind
, we can calculate necessary bounds for the detector efficiency and background levels . for the quantum protocol to beat the best classical one for @xmath21
we need a success rate higher than @xmath23 .
first i will analyse the case of no background detections , and detectors that either obtain the correct result or fail to register the event @xcite . with a finite detector efficiency
@xmath74 , alice and bob must agree on a procedure for the case when their detectors fail .
they are not allowed to communicate the failure , as this would consist of further bits of communication between the parties , and the communication allowed is restricted to one bit .
the most effective way is for alice to proceed with the best classical protocol in case her detector fails , and for bob to do the same when his detector fails .
whenever both detectors fail , alice and bob will obtain the best classical sucess rate . in the case of a single detector failure ,
bob s guess will still be correct with @xmath75 . assuming independent errors at the two detectors , we can calculate the minimum detector efficiency needed : @xmath76 for the case @xmath21 that we analysed , we have @xmath77 and @xmath23 , which results in a minimum necessary detector efficiency @xmath78 which is also the minimum detector efficiency necessary for a loophole - free chsh inequality test @xcite .
a similar analysis can be done for general @xmath7 , and it turns out that the minimum detector efficiency required increases with increasing @xmath7 , making the @xmath21 case the most interesting for an experimental test .
we can include background detections in the analysis by introducing a factor @xmath79 , defined such that @xmath80 is the fraction of detections that are due to background photons .
we assume completely random background detections , yielding each measurement outcome with probability @xmath75 .
note that imperfections in the state preparation procedure can also be incorporated in the model through the parameter @xmath79 .
each run of the experiment can then be classified into one of three categories : \1 ) alice and bob measure their parts of the epr pair accurately with probability @xmath81 , in which case they apply the quantum protocol and succeed with probability @xmath72 .
\2 ) both their detectors fail to detect anything , which will happen in a fraction @xmath82 of the runs .
in this situation they will both use the classical protocol with a probability of success @xmath83 .
\3 ) in all the other situations bob will use either the quantum or the classical protocol , depending on whether he detects a photon or not .
however , due to either background or lack of detection at alice s side his guesses will be random , yielding a success rate of only @xmath29 .
taking all this into account , the condition for a higher - than - classical success rate is : @xmath84 with @xmath83 and @xmath72 from our task with @xmath21 , this is equivalent to @xmath85 in particular , for perfect detectors ( @xmath86 ) , we need @xmath87 . in figure 1
we represent the region in the @xmath88 parameter space that guarantees a higher - than - classical success rate .
the favourable @xmath88 region represents exactly the same conditions necessary for a violation of the chsh inequality ( as can be gathered from @xcite , for example ) .
this is not a surprise , given the equivalence between the quantum protocol and a chsh test pointed out in section [ sec bell ] .
more generally , whenever we have a quantum cc protocol outperforming the optimal classical protocol for a given cc task , the measurements involved in the quantum protocol can be re - interpreted in terms of a non - locality proof .
this is the case at least for tasks allowing only one round of communication between the parties , as the one presented above and the one we will analyse in the next section . for this class of cc tasks , it is always possible to carry out the protocol in such a way as to enforce that all measurements on entangled states be performed in space - like separated regions ; if despite this , the cc task is performed with higher than classical efficiency , it must be because of stronger - than - classical correlations among the results of the measurements of the quantum protocol .
maximally entangled pairs of photons can be generated with high fidelity and measured with precision , which guarantees the equivalent of a very high @xmath79 .
however , detector efficiencies @xmath89 are still short of those necessary . ion traps techniques , on the other hand , can reach high @xmath89 and @xmath79 @xcite .
this makes ion traps a natural choice for the experimental demonstration of the simple quantum communication complexity task presented here .
in this section i will discuss the three - party communication complexity ( cc ) task described in @xcite , finding the necessary experimental requirements for the demonstration of a quantum protocol for it .
the enhancement in performance with relation to the best classical protocol arises from measurements on a tripartite greeberger - horne - zeilinger ( ghz ) state @xcite @xmath90 the state preparation and measurement tend to be much harder to perform experimentally than for the two - party protocol described in section [ sec quant prot ] . in principle , however , we will see that the three - party quantum protocol requires lower detector efficiencies and tolerates higher background count rates than the two - party protocol discussed above .
the cc task is defined as follows @xcite .
alice , bob and claire receive respectively the numbers @xmath1 , @xmath2 and @xmath31 @xmath91 , and these are guaranteed to satisfy @xmath92 the goal is for alice to learn the value of the function @xmath93\ ] ] ( which can be either @xmath5 or @xmath94 ) , after receiving a single bit of communication from each bob and claire .
we assume an uniform probability distribution for @xmath1 , @xmath2 and @xmath31 , subject to satisfying eq .
( [ cond gmn ] ) .
there is a quantum protocol which succeeds with probability @xmath95 ( see @xcite ) .
it is inspired by the ghz non - locality proof of @xcite .
the best classical protocol , however , succeeds only with a lower probability @xmath83 . in order to calculate it , we first note that for each possible @xmath1 all four possibilities for @xmath2 and @xmath31 are allowed by condition ( [ cond gmn ] ) , provided they are properly paired .
this means that by knowing @xmath1 , alice can not infer any information about the values of bob s @xmath2 and claire s @xmath31 , as they are all equally likely candidates .
therefore the best possible procedure for bob will be to agree beforehand on the encoding scheme for his message , so that his bit of communication informs alice that @xmath2 lies in one of two disjunct sets , each containing two possible @xmath2 values .
the same holds for claire and her one - bit message about @xmath31 .
the task involves the assumption that all @xmath96 and @xmath31 values satisfying ( [ cond gmn ] ) will be chosen with equal probabilities . as all @xmath97 possible combinations of @xmath2 and @xmath31 appear the same number of times in the full list of possibilities
, this means that the choice of two - element disjunct sets for the encoding is not important , all yielding the same probability of success @xmath83 .
we can calculate @xmath83 by choosing any such encoding for bob and claire , and averaging over the probability of success for all possible @xmath96 and @xmath31 .
this results in the highest classical success probability @xmath98 if we are to implement the quantum protocol in the laboratory , then we have to account for imperfect detectors and state preparation .
the analysis is similar to that presented in section [ sec experim ] , with the difference that here we have three detectors instead of two .
we define the background rate @xmath80 as the fraction of events detected at each detector which are due to noise ; we assume the worst - case scenario of completely random detection outcomes from these events . the single detector efficiency @xmath89 is the fraction of all events that result in a detector firing .
whenever we detect a signal from the three detectors ( which will happen with probability @xmath99 ) , the quantum protocol works with probability @xmath95 .
if alice s detectors does not fire , she will rely on the possibility of bob s and claire s detectors not firing either , and on them using the optimal classical protocol ; this will happen with probability @xmath100 and result in a probability of success @xmath23 . in all other cases ,
alice will only be able to make random guesses , succeeding with probability @xmath29 .
assuming independent errors at each detector , for a higher - than - classical probability of success we need @xmath101\frac{1}{2}>p_{c}. \label{ineq tri}%\ ] ] substituting @xmath23 and @xmath95 into the inequality above , we obtain @xmath102 this inequality gives us the region in the @xmath103 parameter space that allows for a successful experimental test of the quantum protocol . for perfect detectors ( @xmath86 ) we need @xmath104 whereas for zero background and perfect state preparation and measurement ( @xmath105 ) , it is sufficient to have @xmath106 note that these are less demanding requirements than those necessary for the bipartite quantum cc protocol [ see eqs .
( [ mu and eta ] ) and ( [ bound visibility ] ) ] .
this is related to the fact that quantum non - locality tests for multiparticle states can be done with lower detector efficiencies than for the two - particle case @xcite .
given the possibility of interpreting the higher - than - classical performance of the quantum protocol in terms of a quantum non - locality test ( as was pointed out at the end of section [ sec experim ] ) , we can view requirement ( [ mueta tri ] ) as a sufficient condition for testing quantum non - locality for three maximally entangled particles .
besides limited detector efficiency , here we consider also limited visibility ( due to background counts ) , a situation which has not been considered before .
however , the bound ( [ mueta tri ] ) is not strict ; the simplest way to see this is by noting that with @xmath105 , a ghz test can be performed whenever @xmath107 if we assume independent errors @xcite .
i have presented a simple two - party communication complexity task which can be implemented with a higher - than - classical success rate using quantum entanglement .
the efficiency of the quantum protocol was shown to be linked to its equivalence to measurements that maximally violate the chsh inequality . in order to obtain a higher success rate than classically possible , it is sufficient to have detector efficiencies and background count rates compatible with those necessary for a chsh test ( see fig .
i have also analysed the experimental requirements for a three - party quantum communication complexity protocol presented in @xcite and have shown that the requirements are less demanding than for the two - party protocol discussed .
at least for the two - party protocol , the experimental requirements are within reach of currently available ion trap technology , indicating the feasibility of demonstrating quantum communication complexity protocols experimentally in the near future .
some relations were pointed out between quantum non - locality proofs and quantum communication complexity protocols . in particular , i have shown that sufficient conditions for violation of quantum non - locality inequalities can be derived from the analysis of quantum communication complexity protocols of two or more parties .
i would like to thank jan - ke larsson for pointing out a mistake in an earlier version of this paper and lucien hardy for valuable discussions .
i acknowledge support from the u.k .
overseas research studentships scheme and from the brazilian agency coordenao de aperfeioamento de pessoal de nivel superior ( capes ) . c. h. bennett and g. brassard `` quantum cryptography : public key distribution and coin tossing '' , proceedings of ieee international conference on computer systems and signal processing , bangalore , india , december 1984 , pp 175 - 179 ; d. deutsch _
et al . _ ,
lett . * 77 * , 2818 ( 1996 ) ; _ erratum _ * 80 * , 2022 ( 1998 ) ; d. mayers , lanl e - print quant - ph/9802025 ; h. -k .
lo , h. f. chau , lanl e - print quant - ph/9803006 .
q. a. turchette _
et al . _ ,
lett . * 81 * , 3631 ( 1998 ) ; c. monroe _ et al .
_ , ` scalable entanglement of trapped ions ' , to appear in proc .
17th int . conf . on atomic physics ; d. wineland , private communication ( 2000 ) . | i present a simple two - party quantum communication complexity protocol with higher success rate than the best possible classical protocol for the same task .
the quantum protocol is shown to be equivalent to a quantum non - locality test , except that it is not necessary to close the locality loophole .
i derive bounds for the detector efficiency and background count rates necessary for an experimental implementation and show that they are close to what can be currently achieved using ion trap technology .
i also analyze the requirements for a three - party protocol and show that they are less demanding than those for the two - party protocol .
the results can be interpreted as sufficient experimental conditions for quantum non - locality tests using two or three entangled qubits . |
_ gaia _ s radial velocity spectrometer @xcite data are processed on - ground by the data processing and analysis consortium ( dpac ) co - ordination unit ( cu ) 6 spectroscopic processing pipeline @xcite .
the pipeline formally runs at the cu6 data processing centre .
we present offline tests of the cu6 pipeline running at the mullard space science laboratory . as already presented in @xcite , fig .
[ fig1 ] ( left ) presents tentative evidence that the cu6 pipeline is able to achieve the end - of - mission rv precision predicted by simulations that include the straylight . the majority of the stars in fig .
[ fig1 ] ( left ) are g dwarfs so the measured @xmath4 km s@xmath0 at @xmath2 mag is consistent with simulations that predict a loss of 1.4 mag .
the fixed limiting magnitude of @xmath5 mag onboard rvs was changed to an adaptive one in september 2015 .
now the limiting magnitude ranges from @xmath6 mag as a function of time , corresponding to the straylight pattern to save telemetry on spectra that contain more noise than signal . selecting milky way halo stars from the gums simulation @xcite with @xmath7 mag
suggests that rvs is collecting spectra for @xmath31.8 million stars in the milky way s halo , the majority of which will be metal - poor giants .
having approximately verified the rv performance model in section [ sec1 ] with preliminary measurements , we apply this model to each gums star .
this predicts an _ upper _ limit ( due to the adaptive limiting magnitude ) on the number of halo stars in the final _ gaia _ catalogue with @xmath8 km s@xmath0 : @xmath3800,000 , @xmath342% of the observed halo stars ( red / dark grey dots in fig .
[ fig1 ] right ) .
[ fig1 ] ( left ) presents tentative evidence that @xmath9 can be measured up to 30 km s@xmath0 : @xmath31.2 million stars , 67% of the observed halo stars ( green / light grey dots in fig .
[ fig1 ] right ) . beyond this limit
, it has not been tested that the data follow poisson statistics and so whether the predictions are valid .
rvs spectra were obtained with a fixed across scan ( ac ) window width of 10 pixels .
the rvs s / n simulator suggests that adapting the ac width to the observing conditions could mitigate the impact of straylight , recovering @xmath30.14 mag .
this would increase the aforementioned predicted upper limit to @xmath3900,000 , a 13% increase ( all galactic components increase by several million ) .
_ gaia _ s onboard software was updated in april 2015 to observe with adaptive ac widths .
this new functionality has been successfully commissioned but depends very much on rvs calibrations and so is still being optimised .
mag looking face on : x and y are in the galactic plane with the sun at the origin ( square ) , where the cross is the galactic centre .
each star is colour - coded according to its predicted end - of - mission @xmath9 , assuming every star is a metal - poor k1 giant : @xmath8 km s@xmath0 ( red / dark grey ) , @xmath10 km s@xmath0 ( green / light grey ) , @xmath11 km s@xmath0 ( black).,title="fig:",scaledwidth=49.0% ] mag looking face on : x and y are in the galactic plane with the sun at the origin ( square ) , where the cross is the galactic centre .
each star is colour - coded according to its predicted end - of - mission @xmath9 , assuming every star is a metal - poor k1 giant : @xmath8 km s@xmath0 ( red / dark grey ) , @xmath10 km s@xmath0 ( green / light grey ) , @xmath11 km s@xmath0 ( black).,title="fig:",scaledwidth=41.0% ] _ gaia_-rvs is already the largest ever spectroscopic survey ( 5.4 billion spectra observed in its first year ) .
this means it will also be the largest ever survey of the milky way s halo .
it will provide @xmath31 km s@xmath0 precision radial velocities for @xmath12 mag , which is the planned cu6 contribution to the second _ gaia _ data release ( 2017 , to be confirmed ) . with _
astrometry , this will provide @xmath310,000 _ local field _ halo stars ( all galactic components @xmath32 million ) in 6d .
rvs spectra will also be used by dpac cu8 to derive abundances ( fe , ca , ti , si ) , @xmath13 , log @xmath14 , [ m / h ] and @xmath15 for @xmath12 mag , ready for later data releases .
cu6-measured @xmath16sin@xmath17 for these stars means that _ gaia _ will measure these @xmath310,000 local field halo stars and the @xmath32 million all - galactic - component stars in a total of 15 dimensions .
we present preliminary evidence that _ gaia_-rvs can provide @xmath315 km s@xmath0 precision end - of - mission radial velocities at @xmath2 mag . at this precision and with _ gaia _ astrometry , _ gaia _ will provide @xmath3800,000 - 900,000 halo stars ( all galactic components @xmath375 - 100 million ) in 6d in the final _ gaia _ catalogue ( 2022 , to be decided ) . when analysed with cu8-derived [ m / h ] from _ gaia
_ s bp / rp spectra of these stars , _ gaia _
s chemo-6d - kinematic mapping out to @xmath330 - 50 kpc from the sun will revolutionise our understanding of the milky way halo s structure , origin and evolution . | _ gaia _ s radial velocity spectrometer ( rvs ) has been operating in routine phase for over one year since initial commissioning
. rvs continues to work well but the higher than expected levels of straylight reduce the limiting magnitude .
the end - of - mission radial - velocity ( rv ) performance requirement for g2v stars was 15 km s@xmath0 at @xmath1 mag .
instead , 15 km s@xmath0 precision is achieved at @xmath2 mag , consistent with simulations that predict a loss of 1.4 mag
. simulations also suggest that changes to _ gaia _ s onboard software could recover @xmath30.14 mag of this loss . consequently _ gaia _ s onboard software was upgraded in april 2015 .
the status of this new commissioning period is presented , as well as the latest scientific performance of the on - ground processing of rvs spectra .
we illustrate the implications of the rvs limiting magnitude on _ gaia _ s view of the milky way s halo in 6d using the _ gaia _ universe model snapshot ( gums ) . |
one of the basic quality tests for any model of composite baryons is the level of reached in describing the nucleon and @xmath6 excitation spectra . in that respect ,
the knowledge on the degeneracy group of baryon spectra appears as a key tool in constructing the underlying hamiltonian of the strong - interaction dynamics as a function of the casimir operators of the symmetry group . to uncover the latter
, one can analyze isospin by isospin how the masses of the resonances from the full baryon listing in ref .
@xcite spread with spin and parity .
such an analysis has been performed in prior work @xcite where it was found that breit - wigner masses reveal on the mass / spin ( @xmath7 ) plane a well pronounced spin- and parity clustering .
there , it was further shown that the quantum numbers of the resonances belonging to a particular cluster fit into o(1,3 ) lorentz group representations of the so called rarita - schwinger ( rs ) type @xcite @xmath8\ , .
\label{rs_fields}\ ] ] to be specific , one finds the three rs clusters with @xmath1 , and @xmath2 in both the nucleon @xmath9 and @xmath10 spectra .
as long as the _ lorentz group is locally isomorphic to @xmath5 _ , multiplets with the quantum numbers of the rs representations also appear in typical @xmath5 problems such as the levels of an electron with spin in the hydrogen atom . there ,
the principal quantum number of the coulomb problem is associated with @xmath11 while the rle of the boost generators is taken by the components of the runge - lenz vector .
the rarita - schwinger fields are the so - called `` diagonal case '' ( i.e. @xmath12 ) of the more general representations @xmath13 $ ] .
the rs fields are described in terms of totally symmetric traceless rank-@xmath14 lorentz tensors with dirac spinor components that satisfy the dirac equation for each lorentz index , @xmath15 , associated with a four - vector @xmath16 space @xmath17 the fields of the type in eq .
( [ rs_fields ] ) were considered six decades earlier by rarita and schwinger @xcite , the most popular being the @xmath18-field that has been frequently applied to the description of spin-3/2 particles .
around mid sixties , weinberg @xcite continued the tradition of the original rarita - schwinger work @xcite and considered @xmath19 as fields suited for the description of _ pure spin_-@xmath20 states of _ fixed parity_. the conjecture that @xmath19 can be reduced to a single - spin state was based upon the belief that its lower - spin components are redundant , unphysical states which can be removed by means of the two auxiliary conditions @xmath21 , and @xmath22 . that these conditions do not serve the above purpose ,
was demonstrated in ref .
there , the first auxiliary condition was shown to solely test consistency with the mass - shell relation @xmath23 , while the second one amounted to the acausal energy - momentum dispersion relation @xmath24 .
it is that type of acausality that must be at the heart of the velo - zwanziger problem @xcite .
the rs fields in @xmath5 are in fact compilations of fermions of different spins and parities . to illustrate this statement , and for the sake of concreteness , we here consider the coupling of , say , a positive parity dirac fermion to the @xmath25 hyper - boson the latter being composed of o(3 ) states of either natural @xmath26 , or , unnatural @xmath27 ) parities .
these ( mass degenerate ) @xmath28 states carry all integer internal angular momenta , @xmath29 , with @xmath30 and transform ( for the odd @xmath14 s of interest ) with respect to the space inversion operation @xmath31 according to @xmath32 in coupling now the dirac spinor to @xmath25 from above , the following spin ( @xmath33 ) and parity @xmath34 ) quantum numbers are created @xmath35 in the following , we will use for the spin - sequence in eq .
( [ coupl_scheme ] ) the short - hand notation @xmath36 , with @xmath37 , or , equivalently @xmath38 here , @xmath39 stands for the isospin spinor attributed to the states under consideration .
a glance at the baryon spectra teaches us that actually nature strongly favors the excitations of multi - spin - valued resonance clusters over that of pure higher - spin states .
this circumstance suggests a new data supported interpretation of the rs fields as complete resonance packages . in terms of the notations introduced above , all reported light - quark baryons with masses 2500 mev ( up to the @xmath40 resonance that is most probably an independent quark - gluon hybrid state @xcite ) , have been shown in ref
@xcite to be completely accommodated by the rs clusters @xmath41 , @xmath42 , and @xmath43 , having states of highest spin-@xmath44 , @xmath45 , and @xmath46 , respectively ( see fig . 1 ) .
in each one of the nucleon , @xmath10 , and @xmath47 hyperon spectra , the natural parity cluster @xmath41 is always of lowest mass .
we consider it to reside in a fock space , @xmath48 , built on top of a scalar vacuum .
equations ( [ parity ] ) and ( [ coupl_scheme ] ) illustrate how the @xmath49 clusters ( with @xmath50 , and @xmath51 ) always unite the first spin-@xmath52 , @xmath53 , and @xmath54 resonances . for the non - strange baryons , @xmath49 is followed by the unnatural parity clusters @xmath42 , and @xmath43 , which we view to reside in a different fock space , @xmath55 , built on top of a pseudoscalar vacuum that is orthogonal ( for an ideal @xmath5 symmetry ) to the previous scalar vacuum . to be specific , one finds all the seven @xmath10-baryon resonances @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath60 , @xmath61 and @xmath62 from @xmath63 to be squeezed within the narrow mass region from 1900 mev to 1950 mev , while the @xmath64 resonances paralleling them , of which only the @xmath65 state is still `` missing '' from the data , are located around 1700@xmath66 mev ( see left fig . 1 ) .
therefore , the f@xmath67 resonance is the only non - strange state with a mass below 2000 mev which is `` missing '' for the completeness of the present rs classification scheme . in
further paralleling baryons from the third nucleon and @xmath10 clusters with @xmath68 , one finds in addition the four states @xmath69 , @xmath57 , @xmath58 , and @xmath59 with masses above 2000 mev to be `` missing '' for the completeness of the new classification scheme .
the @xmath69 state is needed to parallel the well established @xmath70 baryon , while the @xmath10-states @xmath57 , @xmath58 , and @xmath59 are required as partners to the ( less established ) @xmath71(2100 ) , @xmath72(1900 ) , and @xmath73(2080 ) nucleon resonances . for @xmath74 hyperons
, incomplete data prevent a conclusive analysis .
even so , fig .
2 ( left ) indicates that the rs motif may already show up in the reported spectrum .
the ( approximate ) degeneracy group of baryon spectra as already suggested in refs .
@xcite , is , therefore , confirmed to be @xmath75 i.e. , isospin@xmath76space - time symmetry .
to summarize , we here state the principle that light unflavored baryon excitations are patterned after lorentz - multiplets .
for example , the rarita
schwinger spinors @xmath77 with @xmath1 , and @xmath2 accommodate all the @xmath78 resonances according to : @xmath79 occasionally , the above structures will be referred to as lampf clusters to emphasize their close relationship to lampf physics . 5.0 cm [ fig : rs_degeneracy ] the scalar vacuum in the first fock space reflects the nambu - goldstone mode of chiral symmetry near the ground state . as argued in ref .
@xcite , its change to a pseudoscalar between the 1st and 2nd clusters , may be related to a change of the mode of chiral symmetry realization in baryonic spectra .
5.0 cm [ fig : o4_band ] within our scheme , the inter - cluster spacing of @xmath80 to @xmath81 mev is larger by a factor of @xmath82 to @xmath83 as compared to the mass spread within the clusters .
for example , the @xmath84 , @xmath85 , @xmath86 , and @xmath63 clusters carry the maximal internal mass splitting of @xmath87 to @xmath88 mev .
finally , the reported mass averages of the resonances from the rs multiplets with @xmath1 , and @xmath2 are well described by means of the following simple _ empirical _ relation : @xmath89 where , again , @xmath90 .
the parameters take for the nucleon @xmath91 ) the values @xmath92 mev , @xmath93 mev , and @xmath94 , respectively . the @xmath6 spectrum
( @xmath95 is best fitted by the smaller @xmath96 value of @xmath97 mev and @xmath98 ( right fig . 2 ) .
it is the goal of this paper to develop a constituent model for baryons that explains the observed clustering in the spectra of the light unflavored baryons .
the paper is organized as follows . in section
ii we motivate legitimacy of fundamental fields of specified mass and unspecified spin as they emerge in the decomposition of a triple - dirac - fermion system into lorentz group representations . in section iii we present the quark version of the diatomic rovibron model @xcite and study its excitation modes .
there we also establish correspondence between excited rovibron states and the baryonic rs clusters .
we further make all the observed and some of the `` missing '' resonances distinguishable in organizing them into different rovibron modes .
we construct the relevant quark hamiltonian and recover eq .
( [ balmer_ser ] ) .
we finally outline the construction of electric transition operators and calculate selected electric transitions of cluster inhabitants to the nucleon .
the paper is finished by a brief summary and outlook .
the relativistic description of three - dirac - spinor systems has been studied in detail in ref .
starting with the well known lorentz invariance of the ordinary dirac equation @xmath99 the authors show that the direct product of three dirac spinors gives rise to a 64-dimensional linear equation of the type @xmath100 here , @xmath101 stays for the four dimensional unit matrix , while the index @xmath102 indicates position of the dirac matrix @xmath103 in @xmath104 . under lorentz transformations ( @xmath105 ) of the @xmath106 matrices , the matrices @xmath107 from eq .
( [ bhabha ] ) change according to @xmath108 = @xmath109 with @xmath110 , and @xmath111 defined as the matrix that covers the lorentz transformation @xmath112=@xmath113 = @xmath114 of @xmath115 . equation ( [ bhabha ] ) is therefore lorentz invariant .
moreover , it was demonstrated that eq .
( [ bhabha ] ) has @xmath3 as an additional dynamical symmetry .
the 64 states from above are distributed over different irreducible representations ( irreps ) of @xmath3 and the permutational group group @xmath116 as well . to be specific
, one finds two 20plets in turn associated with the young schemes @xmath117 $ ] , and @xmath118 $ ] .
they are completed by the quartet @xmath119 $ ] .
the three - dirac spinor state ( denoted by @xmath120 ) can be characterized by the set of quantum numbers @xmath121 here , @xmath122 stands for a set of quantum numbers characterizing the @xmath3 basis vectors of the @xmath123 irrep , while @xmath124 denotes the yamanouchi symbol labeling the basis vectors of the @xmath116 representation @xmath125 @xcite .
the yamanouchi symbols for the @xmath126 , @xmath127 , and @xmath128 are 1 ; 2 , 1 , and 1 , respectively .
the complete number @xmath129 ) of 64 states of the three - dirac - fermion ( @xmath120 ) system is then encoded by the relation @xmath130 where dim@xmath131 and dim@xmath132 are in turn the dimensionalities of the @xmath3 irrep @xmath131 , and the @xmath116 irrep @xmath132 , respectively . in considering now the reduction chain @xmath133 ,
allows for a more detailed specification of the spin content of the @xmath3 multiplets from above ( see ref .
@xcite for details ) .
the quantum numbers of the irreducible representations ( irreps ) of @xmath134 are labeled by the two numbers @xmath135 which can be either integer , or half - integer .
the states participating a given @xmath134 irrep can be further specified by the quantum numbers of the irreps of the @xmath134 subgroups appearing in the reduction chain @xmath136 . to specify the @xmath5 irreps in the context of the @xmath134 reduction down to @xmath137
it is more convenient to use instead of the pair @xmath138 from above , rather the pair @xmath139 with the mapping @xmath140 finally , the @xmath28 irreps in the @xmath141 reduction scheme are labeled by the well known spin ( j ) and magnetic quantum number ( m ) .
the complete set of quantum numbers specifying a member of a @xmath134 multiplet @xmath142 satisfy the inequalities @xmath143 the @xmath3 irrep @xmath127 is of particular interest for the present work . in the @xmath133 reduction chain it splits into @xmath134 irreps according to @xmath144
the first irrep on the rhs of the last equation is 16-dimensional , while the second is four dimensional and associated with a dirac spinor . as we shall see below
, the @xmath134 16plet @xmath145 is nothing but the rs field with @xmath146 . indeed , from eq .
( [ schachtelung ] ) follows that @xmath147 the inequalities in the latter equation are satisfied for @xmath148 , @xmath149 , and for @xmath150 , @xmath151 . in accordance with the 2nd equation in ( [ schachtelung ] )
, @xmath33 can take the three values @xmath152 , and @xmath153 .
thus , the @xmath154 irrep of o(5 ) describes a spin-3/2 and two spin-1/2 states and coincides with the lowest 16-dimensional rarita - schwinger field .
the above consideration gives an idea of how lorentz representations of the rs type can emerge as _ fundamental _ free particles of definite mass and indefinite spin within the context of a relativistic space - time treatment .
though such point - like particles have not been detected so far , the @xmath155 and @xmath6 spectra strongly indicate existence of _ composite _ rs fields . in the following ,
we shall focus onto that very realization of multi - spin lorentz representations and explore their internal structure by means of constituent models . for a more profound textbook presentation on the various aspects of higher - dimensional relativistic supermultiplets ,
the interested reader is referred to ref .
baryons in the quark model are considered as constituted of three quarks in a color singlet state .
it appears naturally , therefore , to undertake an attempt of describing the baryonic system by means of algebraic models developed for the purposes of triatomic molecules , a path already pursued by refs .
@xcite . there ,
the three body system was described in terms of two vectorial ( @xmath156 ) and one scalar @xmath157 boson degrees of freedom that transform as the fundamental @xmath158 septet . in the dynamical symmetry limit @xmath159 the degrees of freedom associated with the one vectorial boson factorize from those associated with the scalar boson and the remaining vectorial boson .
because of that the physical states constructed within the @xmath158 ibm model are often labeled by means of @xmath160 quantum numbers
. below we will focus on that very sub - model of the ibm and show that it perfectly accommodates the rs clusters from above and thereby the lampf data on the non - strange baryon resonances .
the dynamical limit @xmath161 corresponds to the quark
diquark approximation of the three quark system , when two of the quarks reveal a stronger pair correlation to a diquark ( dq ) @xcite , while the third quark ( q ) acts as a spectator .
the diquark approximation turned out to be rather convenient in particular in describing various properties of the ground state baryons @xcite . within the context of the quark
diquark ( q - dq ) model , the ideas of the rovibron model , known from the spectroscopy of diatomic molecules @xcite , can be applied to the description of the rotational - vibrational ( rovibron ) excitations of the q dq system . in the rovibron model ( rvm ) the relative q dq motion ( see fig .
3 ) is described by means of four types of boson creation operators @xmath162 , and @xmath163 ( compare @xcite ) . the operators @xmath164 and @xmath165 in turn transform as rank-0 , and rank-1 spherical tensors , i.e. the magnetic quantum number @xmath166 takes in turn the values @xmath167 , @xmath51 , and @xmath168 . in order to construct boson - annihilation operators that also transform as spherical tensors ,
one introduces the four operators @xmath169 , and @xmath170 .
constructing rank-@xmath171 tensor product of any rank-@xmath172 and rank-@xmath173 tensors , say , @xmath174 and @xmath175 , is standard and given by @xmath176 here , @xmath177 are the well known @xmath28 clebsch - gordan coefficients . [
fig : ro_vibron ] now , the lowest states of the two - body system are identified with @xmath178 boson states and are characterized by the ket - vectors @xmath179 ( or , a linear combination of them ) within a properly defined fock space .
the constant @xmath180 stands for the total number of @xmath181- and @xmath182 bosons and plays the rle of a parameter of the theory . in molecular physics , the parameter @xmath155 is usually associated with the number of molecular bound states .
the group symmetry of the rovibron model is well known to be @xmath3 .
the fifteen generators of the associated @xmath183 algebra are determined as the following set of bilinears @xmath184 the @xmath185 algebra is then recovered by the following commutation relations @xmath186 the operators associated with physical observables can then be expressed as combinations of the @xmath185 generators . to be specific
, the three - dimensional angular momentum takes the form @xmath187 further operators are @xmath188) and @xmath189 ) defined as @xmath190 respectively . here
, @xmath191 plays the rle of the electric dipole operator .
finally , a quadrupole operator @xmath192 can be constructed as @xmath193 the @xmath185 algebra has the two algebras @xmath194 , and @xmath195 , as respective sub - algebras . the @xmath194 algebra is constituted by the three generators @xmath196 , and the five components of the quadrupole operator @xmath192
. its @xmath195 subalgebra is constituted by the three components of the angular momentum operator @xmath196 , on the one side , and the three components of the operator @xmath197 , on the other side .
thus there are two exactly soluble rvm limits that correspond to the two different chains of reducing @xmath3 down to @xmath28 .
these are : @xmath198 respectively .
the hamiltonian of the rvm in these exactly soluble limits is then constructed as a properly chosen function of the casimir operators of the algebras of either the first , or the second chain . for example , in case one approaches @xmath28 via @xmath199 , the hamiltonian of a dynamical @xmath200 symmetry can be cast into the form : @xmath201 here , @xmath202 is a constant , @xmath203 , and @xmath204 are in turn the quadratic ( in terms of the generators ) casimirs of the groups @xmath200 , and @xmath205 , respectively , while @xmath206 and @xmath207 are constants , to be determined from data fits .
a similar expression ( in obvious notations ) can be written for the rvm hamiltonian in the @xmath208 exactly soluble limit : @xmath209 the casimir operator @xmath210 is defined accordingly as @xmath211 and has an eigenvalue of @xmath212 . in molecular physics , only linear combinations of the casimir operators are used , as a rule . however , as known from the hydrogen atom @xcite , the hamiltonian is determined by the inverse power of @xmath213 according to @xmath214 where @xmath215 is a parameter with the dimensionality of mass .
this hamiltonian predicts the energy of the states as @xmath216 and does not follow the simple linear pattern ( see also eq .
( [ h_so4 ] ) ) . in order to demonstrate how the rvm applies to baryon spectroscopy ,
let us consider the case of q - dq states associated with @xmath217 and for the case of a @xmath218 dynamical symmetry . from now on we shall refer to the quark rovibron model as qrvm .
it is of common knowledge that the totally symmetric irreps of the @xmath185 algebra with the young scheme @xmath219 contain the @xmath218 irreps @xmath220 with @xmath221 each one of these @xmath218 irreps contains @xmath205 multiplets with three dimensional angular momenta @xmath222 in applying the branching rules in eqs .
( [ sprung_k ] ) , ( [ o3_states ] ) to the case @xmath217 , one encounters the series of levels @xmath223 the parity carried by these levels is @xmath224 where @xmath225 is the parity of the relevant vacuum . in coupling
now the angular momenta in eq .
( [ ns_ks ] ) to the spin-1/2 of the three quarks in the nucleon , the following sequence of states is obtained : @xmath226 thus rovibron states of half - integer spin will transform according to @xmath227 $ ] representations of @xmath218 .
the isospin structure is accounted for pragmatically through attaching to the rs clusters an isospin spinor @xmath39 with @xmath101 taking the values @xmath228 and @xmath229 for the nucleon , and the @xmath6 states , respectively . as illustrated by fig . 1 ,
the above quantum numbers cover both the nucleon and the @xmath6 excitations .
note that in the present simple version of the rovibron model , the spin of the quark
diquark system is @xmath230 , and the total spin @xmath33 takes the values @xmath231 in accordance with eqs .
( [ ns_ks ] ) and ( [ set_q ] ) .
the strong relevance of _ same _ picture for both the nucleon and the @xmath232 spectra ( where the diquark is in a vector - isovector state ) hints onto the dominance of a scalar diquark for both the excited nucleon and @xmath233 states .
this situation is reminiscent of the @xmath234 configuration of the @xmath235plet of the canonical @xmath236 symmetry where the mixed symmetric / antisymmetric character of the @xmath237 wave function in spin - space is compensated by a mixed symmetric / antisymmetric wave function in coordinate space , while the isotriplet @xmath238 part is totally symmetric .
we here will leave aside the discussion of the generic problem of the various incarnations of the ibm model regarding the symmetry properties of the resonance wave functions to a later date and rather concentrate in the next subsection onto the `` missing '' resonance problem . the comparison of the states in eq .
( [ set_q ] ) with the reported ones in eq .
( [ class_scheme ] ) shows that the predicted sets are in agreement with the characteristics of the non - strange baryon excitations with masses below @xmath239 2500 mev , provided , the parity @xmath225 of the vacuum changes from scalar ( @xmath240 ) for the @xmath18 , to pseudoscalar ( @xmath241 ) for the @xmath242 clusters . a pseudoscalar `` vacuum '' can be modeled in terms of an excited composite diquark carrying an internal angular momentum @xmath243 and maximal spin @xmath244 . in one of the possibilities
the total spin of such a system can be @xmath245 .
to explain the properties of the ground state , one has to consider separately even @xmath155 values , such as , say , @xmath246 . in that case
another branch of excitations , with @xmath247 , @xmath248 , and @xmath51 will emerge .
the @xmath249 value characterizes the ground state , @xmath250 corresponds to @xmath251 , while @xmath247 corresponds to @xmath252 . these are the multiplets that we will associate with the `` missing '' resonances predicted by the rovibron model . in this manner ,
reported and `` missing '' resonances fall apart and populate distinct @xmath3- and @xmath218 representations . in making observed and `` missing '' resonances distinguishable , reasons for their absence or , presence in the spectra are easier to be searched for .
as to the parity of the resonances with even @xmath14 s , there is some ambiguity . as a guidance one may consider the decomposition of the three - quark ( @xmath253 ) hilbert space into lorentz group representations as performed in ref .
there , two states of the type @xmath251 were found .
the first one arose out of the decomposition of the @xmath253-hilbert space spanned by the @xmath254 single - particle states .
it was close to @xmath255 and carried opposite parity to the latter .
it accommodated , therefore , unnatural parity resonances .
the second @xmath250 state was part of the @xmath256- single - particle configuration space and was closer to @xmath257 .
it also carried opposite parity to the latter and accommodated natural parity resonances .
finally , the @xmath247 cluster @xmath258 emerged in the decomposition of the one - particle - one - hole states within the @xmath259 configuration space and carried also natural parity , that is , opposite parity to @xmath260 . in accordance with the above results
, we here will treat the @xmath261 states to be all of natural parities and identify them with the nucleon @xmath262 , the natural parity @xmath250 , and the natural parity @xmath247 rs clusters .
the unnatural parity @xmath250 cluster from @xcite could be generated through an unnatural parity @xmath263 excitation mode .
however , this mode would require manifest chiral symmetry up to @xmath264 mev which contradicts at least present data . with this observation in mind
, we here will restrict ourselves to the consideration of the natural parity @xmath261 clusters . in this manner the unnatural parity @xmath250 state from ref .
@xcite will be dropped out from the current version of the rovibron model . from now on we will refer to the excited @xmath261 states as to `` missing '' rovibron clusters . now , the qrvm hamiltonian that reproduces the mass formula from eq .
( [ balmer_ser ] ) is given by the following function of @xmath210 @xmath265 the states in eq .
( [ set_q ] ) are degenerate and the dynamical symmetry is @xmath218 .
the parameter set @xmath266 with @xmath267 , recovers the empirical mass formula in eq .
( [ balmer_ser ] ) .
thus , the @xmath218 dynamical symmetry limit of the qrvm picture of baryon structure motivates existence of quasi - degenerate clusters of resonances in the nucleon- and @xmath6 baryon spectra . in table
i we list the masses of the rs clusters concluded from eqs .
( [ h_qrvm ] ) , and ( [ f_s ] ) . .
predicted mass distribution of observed ( obs ) , and missing ( miss ) rovibron clusters ( in mev ) according to eq .
( 34,35 ) .
the sign of @xmath225 in eq .
( 3 ) determines natural- ( @xmath268 ) , or , unnatural ( @xmath241 ) parity states .
all @xmath6 excitations have been calculated with @xmath97 mev rather than with the nucleon value of @xmath93 mev .
the experimental mass averages of the resonances from a given rs cluster have been labeled by `` exp '' .
the nucleon and @xmath6 ground state masses @xmath269 and @xmath270 were taken to equal their experimental values . [
cols="<,^,^,^,^,^,^,^ " , ] the data on the @xmath74 , @xmath271 , and @xmath272 hyperon spectra are still far from being as complete as those of the nucleon and the @xmath6 baryons and do not allow , at least at the present stage , a conclusive statement on relevance or irrelevance of the rovibron picture fig . 1 .
the presence of the heavier strange quark can significantly influence the excitation modes of the @xmath253-system . in case ,
the presence of the @xmath181 quark in the hyperon structure is essential , the @xmath273 chain can be favored over @xmath208 and a different clustering motif can appear here . for the time being , this issue will be dropped out of further consideration . in the next subsection , we shall outline the calculational scheme for branching ratios of reduced probabilities for electromagnetic transitions . in the following
, resonance states from a rs cluster will be denoted as @xmath274 here , @xmath275 denotes the parity of the vacuum of the fock space accommodating the rs cluster , @xmath276= @xmath277 , @xmath29 is the underlying three - dimensional angular momentum of parity @xmath278 , @xmath279 is the quark spin , while @xmath280 and @xmath281 are in turn total spin and magnetic quantum numbers of the resonance under consideration . in fact , @xmath14 is nothing but the four - dimensional angular momentum . within the framework of the rovibron model
one can describe three different types of transitions : \(i ) transitions without change of the quantum numbers @xmath155 and @xmath14 , i.e. transitions between resonances from same cluster .
in such a case , the transition operator is the @xmath197 generator of the @xmath195 algebra and one can calculate the reduced probabilites @xmath282 for electric dipole transitions . notice that the reduced transition probability of the multipolarity @xmath283 as carried out by the operator @xmath284 between states of initial and final spins @xmath285 and @xmath286 , respectively , is defined as @xcite @xmath287 unfortunately , such transitions are difficult and perhaps even beyond any possibility of being observed .
\(ii ) transitions between states of same number of bosons @xmath155 but of different four dimensional angular momenta , @xmath288 , i.e. transitions between resonances belonging to different rs clusters .
operators that can realize such transitions between different @xmath5 multiplets are @xmath3 generators ( or , tensor products of them ) lying outside of the @xmath195 sub - algebra .
the latter operators constitute the set @xmath289 it is not difficult to prove that the nine operators in eq .
( [ elm_set ] ) behave with respect to @xmath218 transformation as the components of the totally symmetric rank-2 tensor , @xmath290 where @xmath291 by the way , the tensor @xmath292 is the one of lowest rank that can realize transitions between @xmath218 multiplets having same number of bosons @xmath155 and differing by two units in @xmath14 .
\iii ) transitions between @xmath3 multiplets whose number of bosons differ by one unit ( @xmath293 ) , the most interesting being resonance de - excitation modes into the nucleon @xmath294 in the following we will be mainly interested in transitions of the third type . at the present stage , however , it is convenient to first outline the general scheme of the @xmath218 racah algebra .
tensor products @xmath295^{\left ( a , b\right ) lm } $ ] in @xmath218 are defined as ( see refs .
@xcite for details ) @xmath296^{\left ( a , b \right ) lm } = \sum_{l_1m_1 l_2m_2 } & & \nonumber\\ \big ( \left ( a_1b_1\right ) \ , l_1m_1\ , \left ( a_2b_2\right ) \ , l_2m_2 \big\vert \left ( a_1b_1\right)\ , \left ( a_2b_2\right)\ , ; \left ( a b\right ) lm \big ) & & t^{\left ( a_1,b_1 \right)\ , l_1m_1 } t^{\left ( a_2,b_2 \right)\ , l_2m_2 } \ , .
\label{tensor_pr_o4}\end{aligned}\ ] ] the matrix elements of any tensor operator @xmath297 between @xmath5 states are expressed as @xmath298 the @xmath218 clebsch - gordan coefficients entering the last equation are determined by @xmath299 the last equation shows that the ratios of the reduced probabilities of electromagnetic transitions between resonances with different @xmath14 quantum numbers are determined as ratios of the squared @xmath218 clebsch - gordan coeffecients , as the triple barred transition matrix elements cancel out . as an example of that type of transitions let us consider the electromagnetic de - excitations of the natural parity resonances with spins @xmath44 and @xmath300 from the first cluster to the nucleon .
obviously , the relevant tensor operator in @xmath301 space is @xmath302 .
the latter should connect @xmath3 states with different numbers of bosons i.e. @xmath293 .
therefore , it can be taken in the form @xmath303 transitions of the above type can then be calculated by means of ordinary racah algebra in considering @xmath304 ( with @xmath305 ) as an intrinsic quantum number according to : @xmath306 in order to express double barred matrix element in terms of triple barred matrix elements , the following relations should be taken into account : @xmath307 in combining eqs .
( [ yu_1 ] ) and ( [ yu_2 ] ) results into @xmath308 eqs .
( [ yu_1])-([final_me ] ) can be applied to calculate the ratio of , say , the electric dipole de - excitations d@xmath309(1520)@xmath310 , and s@xmath311(1535)@xmath310 . in this case
@xmath312 , @xmath313 , @xmath314 , and @xmath280 takes the two values @xmath315 , and @xmath53 , respectively .
substitution of the relevant quantum numbers into eqs .
( [ yu_1])-([yu_2 ] ) followed by a calculation of the ratio of the squared values of the @xmath315 , and @xmath316 matrix elements yields the theoretical ratio of the electric dipole widths of interest , @xmath317 , and @xmath318 of the respective d@xmath319 and s@xmath320 states as @xmath321 in order to compare it to data , one may approximate the dipole widths with the total @xmath106 widths and obtain their experimental values from the full widths and the branching ratios listed in @xcite .
the full widths of the d@xmath319 and s@xmath320 resonances are reported as 120 mev and 150 mev , respectively .
the d@xmath309(1520)@xmath322 branching ratio is reported as 0.46 - 0.56% , while the s@xmath311(1535 ) takes values within the broader range from 0.15% to 0.35% .
the theoretical prediction corresponds to a s@xmath311(1535)@xmath322 ratio of 0.35% and lies thereby at the upper bound of the data range .
this ratio is in fact @xmath33-independent .
it shows that the purely algebraic description is insufficient to reproduce the electromagnetic properties of the resonances in great detail . in that regard , further development of the model
is needed with the aim to account for the internal diquark structure .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ remarkably , the internal structure of the diquark does not show up in the spectra , and seems to be less relevant for the gross features of the excitation modes . at the vertex level , however , it will gain more importance .
the merit of the rovibron model is that there it can be treated as a correction rather than as a leading mechanism from the very beginning .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ one can further compare gamma - widths of resonances carrying different internal @xmath28 quantum numbers @xmath29 .
this effect is easiest to study on the example of the natural parity resonances from the `` missing '' rovibron clusters . to be specific
, we will compare the reduced probabilities for the following two transitions : @xmath323 the relevant transition operator is @xmath324 here , @xmath29 can take the values @xmath325 , and @xmath326 .
the first of the transitions in eq .
( [ missing_trans ] ) is governed by the electric dipole operator @xmath327 , while the second is controlled by the electric octupole @xmath328 .
we are going to calculate the ratio @xmath329 of the quantities @xmath330 here @xmath331 usage of eq .
( [ final_me ] ) yields equal reduced probabilities for both the dipole and octupole de - excitations and thereby the unit value for @xmath329 .
thus , within this early version of the rovibron model , a given rs cluster will have a common partial ( @xmath332)- decay width , that is insensitive to its @xmath28 spin content .
a more interesting situation occurs in the case of lampf clusters , such like @xmath333 .
there , one encounters a _
suppression of electromagnetic transitions to the nucleon_. indeed , in the rigorous case of an ideal o(4 ) symmetry , due to the unnatural parities of the nucleon resonances with masses above 1535 mev ( and the @xmath6 excitations with masses above 1700 mev ) , transitions of the type @xmath334 can not proceed neither via electric @xmath335- nor via magnetic @xmath336 multipoles ( to be presented elsewhere ) . in the less rigid scenario of a violated @xmath5 symmetry , mixing between states of same parity and total spins but different
@xmath14 s may occur .
for example , the above unnatural parity spin-@xmath54 resonance from the @xmath337 multiplet may mix up with the spin-@xmath54 of natural @xmath338 from the @xmath339 multiplet @xmath340 for similar reasons , also a mixing with @xmath341 states can take place . within this mixing scheme
, unnatural parity resonances can be excited electrically via their natural parity component . as long
as the relevant transition operator for such transitions is @xmath342 , its matrix element between the nucleon and the resonance of interest will be proportional to the mixing parameter @xmath206 . to be specific ,
@xmath343 it is obvious from the last equation , that electric excitations of the nucleon into the unnatural parity resonances will be suppressed by the factor of @xmath344 . at the present early stage of development of the quark rovibron model , the mixing parameter @xmath206
can not be calculated but has to be considered as free and determined from data . a theoretical prediction for @xmath345 would require more fundamental approach to the internal diquark dynamics . in case the @xmath5 symmetry
is slightly violated , one may assume @xmath206 to be same for all cluster inhabitants and perform some calculations as to what extent such states can be linked via electromagnetic transitions to the nucleon .
the results of the present study can be summarized as follows : 1 .
the present investigation communicated an idea of how lorentz representations of the rs type can emerge as _ fundamental _ as well as _ composite _
free particles of definite mass and indefinite spin within the context of a relativistic space - time treatment of the three dirac - fermion system .
though structureless rs particles have not been detected so far , the @xmath155 and @xmath6 spectra strongly indicate existence of _ composite _ rs fields .
excited light unflavored baryons preferably exist as multi - resonance clusters that are described in terms of rs multiplets such as the ( predominantly ) observed lampf clusters @xmath49 , @xmath42 and @xmath43 , and the `` missing '' clusters @xmath346 and @xmath347 .
the above rs clusters accommodate all the resonances observed so far in the @xmath78 decay channel ( up to the @xmath40 state ) .
the lampf data constitute , therefore , an almost accomplished excitation mode in its own rights , as only 5 resonances are `` missing '' for the completeness of this structure .
4 . we modeled composite rs fields within the framework of the quark rovibron model and constructed a hamiltonian that fits the masses of the lampf clusters .
5 . in using that hamiltonian we predicted , from a different but the @xmath236 perspective
, the masses of two `` missing '' clusters of natural parity resonances , in support of the tjnaf `` missing '' resonance search program @xcite .
`` missing '' resonances under debate in the literature , such like p@xmath311(1880)@xcite and p@xmath309(1910)@xcite could neatly fit into the @xmath348 rs cluster at 1935 mev in table i. 6 .
we constructed electric transition operators , outlined the essentials of the o(4 ) racah algebra , and calculated ratios of reduced probabilities of various resonance de - excitations to the nucleon .
we found the internal structure of the diquark to be of minor importance for the gross features of the excitation modes . at the vertex level , however , a point - like diquark was shown to be insufficient to account for differences in the branching ratios of resonances from same cluster .
it is that place where the present early version of the qrvm model of baryon structure needs further improvements . treating the internal structure of the diquark as a correction rather than as a leading mechanism from the very beginning is a major merit of the quark rovibron model .
work supported by conasyt mexico .
m. moshinsky and yu .
f. smirnov , _ the harmonic oscillator in modern physics _ ( harwood academic publishers , n. y. ) 1996
. f. iachello , phys .
lett . * 78 * , 13 ( 1989 ) ; + r. bijker , f. iachello , and a. leviatan , phys .
* c54 * , 1935 ( 1996 ) .
+ r. bijker , f. iachello , and a. leviatan , ann . of phys . *
236 * , 69 ( 1994 ) g. f. filippov , v. i. ovcharenko , and yu . f. smirnov , _ microscopic theory of collective excitations of atomic nuclei _
( naukova dumka , kiev ) 1981 ( in russian ) .
kris l. g. heyde , _ the nuclear shell model _ ( springer verlag , berlin ) . | the structure of the reported excitation spectra of the light unflavored baryons is described in terms of multi - spin valued lorentz group representations of the so called rarita - schwinger ( rs ) type @xmath0 $ ] with @xmath1 , and @xmath2 .
we first motivate legitimacy of such pattern as fundamental fields as they emerge in the decomposition of triple fermion constructs into lorentz representations .
we then study the baryon realization of rs fields as composite systems by means of the quark version of the @xmath3 symmetric diatomic rovibron model . in using the @xmath4 reduction chain , we are able to reproduce quantum numbers and mass splittings of the above resonance assemblies .
we present the essentials of the four dimensional angular momentum algebra and construct electromagnetic tensor operators .
the predictive power of the model is illustrated by ratios of reduced probabilities concerning electric de - excitations of various resonances to the nucleon . |
while transport calculations have , in general , been very successful in describing particle production in heavy - ion collisions @xcite , one of the remaining open problems in the transport - theoretical description of heavy - ion collisions at sis energies ( @xmath3 a gev ) is an overprediction of the pion multiplicity @xcite . at the beam energies of a few a gev the dominating mechanism of the pion production is the excitation of the @xmath4 resonance in a nucleon - nucleon ( nn ) collision followed by its decay : @xmath0 , @xmath5 .
the pion multiplicity , therefore , depends crucially on the value of the in - medium @xmath0 cross section which by detailed balance also determines the pion reabsorption . in ref .
@xcite a phenomenological density - dependent modification factor of the vacuum @xmath0 cross section has been proposed .
the buu calculations with the density dependent @xmath6 cross sections are in a reasonable agreement with the experimental data on pion production @xcite . the purpose of the present work is to get rid of the ad - hoc description of ref .
@xcite and to calculate the in - medium @xmath6 cross sections on the basis of the one - pion exchange model ( opem ) with medium modifications .
the calculated cross sections are used then in the buu model to study the pion production in heavy - ion collisions at sis energies .
it has been shown in ref .
@xcite within the relativistic dirac - brueckner approach that the @xmath0 cross section is reduced at high densities .
later studies within the opem including pion collectivity and vertex corrections @xcite have shown , however , that the @xmath7 cross section grows with nuclear density .
an important contribution which was missed in refs .
@xcite is given by the dirac effective nucleon and @xmath8 resonance masses .
we will show , that the introduction of the dirac effective nucleon and @xmath8 masses in the opem leads to an in - medium reduction of the @xmath6 cross sections in agreement with ref .
@xcite .
the structure of the paper is as follows : in sect .
ii we describe the opem with medium modifications including pion collectivity , vertex renormalization by the contact interactions and effective masses for the nucleon and @xmath8 resonance . the energy dependence of the total @xmath9 cross section as well as the angular and @xmath8 resonance mass dependencies of the differential @xmath9 cross section are studied at finite nuclear matter densities .
iii explains how the in - medium modified cross sections are implemented into the buu code and describes the results of the buu calculations of pion production from heavy - ion collisions at 0.5@xmath102 a gev . in sect .
iv the summary of our results and of remaining problems is given .
for the calculation of the @xmath0 cross section in nuclear matter we apply the nonrelativistic opem similar to ref .
@xcite , however , with readjusted parameters .
we have chosen the nonrelativistic version since it incorporates the contact nuclear interactions in a natural way , which is necessary in the in - medium calculations . in the nonrelativistic reduction ( c.f .
@xcite ) the @xmath11 and @xmath12 interaction lagrangians are : @xmath13 where @xmath14 and @xmath15 are the nucleon , @xmath8 resonance and pion field respectively . @xmath16 and @xmath17 are the coupling constants @xcite .
@xmath18 and @xmath19 are the spin and isospin pauli matrices . @xmath20 and @xmath21 are the spin and isospin transition ( @xmath22 ) operators defined according to ref .
@xcite .
the nuclear spin - isospin short range correlations are known to be important for the in - medium @xmath0 cross section @xcite .
we will introduce them via the following lagrangian : @xmath23 \nonumber \\ & & - { f_\delta^2 \over m_\pi^2 } g_{\delta\delta}^\prime ( \psi^\dag_\delta s_\alpha { \bf t } \psi ) ( \psi^\dag s_\alpha^\dag { \bf t}^\dag \psi_\delta ) \nonumber \\ & & - \left [ { f_\delta^2 \over 2 m_\pi^2 } g_{\delta\delta}^\prime ( \psi^\dag_\delta s_\alpha { \bf t } \psi ) ( \psi^\dag_\delta s_\alpha { \bf t } \psi ) + \mbox{h.c . } \right ] \label{lsrc}\end{aligned}\ ] ] @xmath24 , @xmath25 and @xmath26 are the landau - migdal parameters .
the values of the landau - migdal parameters are not fixed unambiguously in the literature . within a simple universality assumption @xmath27 , which is the bckmann - weise choice ( c.f .
@xcite and refs .
therein ) , one gets @xmath28 from the best description of the unnatural parity isovector states in @xmath29he , @xmath30o and @xmath31ca .
however , the same calculations within the migdal model @xcite assumption @xmath32 produce @xmath33
. the description of the quenching of the gamow - teller matrix elements requires @xmath34 ( assuming @xmath35 ) @xcite . from the real part of the pion optical potential in @xmath36-atoms one gets @xmath37 and @xmath38 @xcite .
the pion induced two - proton emission is the best described with @xmath39 @xcite .
we will adopt the two sets of the landau - migdal parameters from ref .
@xcite ( see table 1 ) . as a first step ,
we calculate the @xmath0 cross section in vacuum .
the in - medium corrections are introduced in a second step .
the differential @xmath40 cross section can be written as : @xmath41 where @xmath42 is the square of the absolute value of the reaction amplitude @xmath43 in the normalization of ref .
@xcite averaged over spins of incoming particles 1,2 and summed over spins of outgoing particles 3,4 .
the flux factor is : @xmath44 .
the spectral function of the @xmath8 resonance @xmath45 where @xmath46 gev is the @xmath8 pole mass , satisfies the normalization condition : @xmath47 the mass dependent total @xmath8 width @xmath48 is parameterized according to ref .
@xcite : @xmath49 where @xmath50 gev is the width at the pole mass , @xmath51 gev / c is the cut - off parameter . in eq.([gamdel ] ) and below @xmath52 is the center - of - mass ( c.m . ) momentum of outgoing particles with masses @xmath53 and @xmath54 from the decay of a particle with mass @xmath55 . in the c.m .
frame of colliding nucleons one gets after standard transformations : @xmath56 where @xmath57 . the amplitude @xmath43 is evaluated from the graphs shown in fig .
[ fig : nnnd_opem ] . in order to preserve the symmetry of the cross section with respect to @xmath58 we have included the contact interaction both in the direct and in the exchange diagrams .
the same was done in refs .
@xcite whereas in ref .
@xcite the contact interaction has been included in the direct term only .
for the discussion on this subject we refer the reader to the appendix of ref .
@xcite . applying feynman rules with the lagrangians ( [ lpinn])-([lsrc ] ) one gets the following expressions for the direct ( a ) and exchange ( b ) terms of the amplitude : @xmath59 id^{(0)}(k ) [ \chi^\dag(3 ) \gamma_j^{dec}(k ) \chi(1 ) ] - iv_{4,3;2,1}(k)~ , \label{ta } \\ it_b & = & -[\chi^\dag_\delta(4 ) \gamma_j^{abs}(k^\prime ) \chi(1 ) ] id^{(0)}(k^\prime ) [ \chi^\dag(3 ) \gamma_j^{dec}(k^\prime ) \chi(2 ) ] + iv_{4,3;1,2}(k^\prime)~. \label{tb}\end{aligned}\ ] ] here @xmath60 and @xmath61 are the spin - isospin wave functions of the nucleon and delta : @xmath62 where @xmath63 ( @xmath64 ) and @xmath65 ( @xmath66 ) are the quantum numbers of the third isospin and spin component of a nucleon ( delta ) .
the quantities @xmath55 , @xmath67 , @xmath68 and @xmath69 are the corresponding indices of the wave functions .
the four - momenta of the exchange pion in the direct and exchange channels are @xmath70 and @xmath71 , respectively .
below we use the kinematical invariants @xmath72 and @xmath73 .
the operators @xmath74 represent the @xmath75 absorption and @xmath76 decay vertices in vacuum .
the quantity @xmath77 [ \chi^\dag(3 ) \sigma_\alpha \mbox{\boldmath${\mathbf\tau}$\unboldmath } \chi(1 ) ] \label{vcont0}\ ] ] is the contact interaction in vacuum and @xmath78 is the free pion propagator .
in the @xmath11 and @xmath12 vertices we use the monopole form factor @xmath79 : @xmath80 with @xmath81 being the cut - off parameter ( see table 1 ) . following refs .
@xcite a factor @xmath82 is included in the contact interaction vertices , since these vertices simulate the exchange by some heavy meson .
the matrix element squared is @xmath83 we now consider the case of the @xmath9 reaction .
the spin - averaged direct term in ( [ tsq ] ) is @xmath84~. \label{tasq}\ ] ] the spin - averaged exchange term @xmath85 and @xmath86 .
the spin - averaged interference term can be written as follows : @xmath87 = - { f_\delta(t)f(t)f_\delta(u)f(u ) \over 2m_\pi^4 } \left [ { 4 \over 3}(({\bf k \cdot k^\prime})^2 + { \bf k}^2 { \bf k^\prime}^2 ) d^{(0)}(k ) d^{(0)}(k^\prime ) \right .
\nonumber \\ & + & \left . { 16 \over 3 } { \bf k}^2 d^{(0)}(k ) g_{n\delta}^\prime + { 16 \over 3 } { \bf k^\prime}^2 d^{(0)}(k^\prime ) g_{n\delta}^\prime + 16(g_{n\delta}^\prime)^2 \right]~. \label{tint}\end{aligned}\ ] ] in numerical calculations we also took into account the c.m .
and relativistic corrections to the amplitudes ( [ ta]),([tb ] ) ( c.f . ref .
@xcite ) : ( i ) we replaced in ( [ gama0 ] ) @xmath88 and @xmath89 in the direct and exchange terms respectively . here
@xmath90 ( @xmath91 ) is the momentum of the exchanged pion in the rest frame of @xmath8 for the direct ( exchange ) term : @xmath92 ( ii ) the direct amplitude @xmath93 is multiplied by the factor @xmath94 and the exchange amplitude @xmath95 by the factor @xmath96 .
this restores approximately the lorentz invariance after summation of the amplitude squared over spin projections . after the c.m . and relativistic corrections eqs.([tasq]),([tint ] )
get slightly modified .
we do not show here explicit formulas : they will be given below for the more general in - medium case ( see eqs.([tamedsq]),([tintmedsq ] ) in appendix a ) .
in the upper panel of fig . [
fig : sigma_vs_plab ] we present the total cross section @xmath97 as a function of the c.m . energy . dashed and
solid lines correspond to the case of set 1 and set 2 , respectively ; the dotted line represents the relativistic calculation using the model of ref .
the relativistic and our corrected nonrelativistic calculations are in a good agreement at @xmath98 gev . at higher c.m .
energies our calculations produce somewhat larger cross section than the relativistic result of ref . @xcite . in the lower panel of fig .
[ fig : sigma_vs_plab ] we show the total cross section of the process @xmath99 vs the c.m .
energy in comparison to the data from ref . @xcite . in this case
the dominating channel is still @xmath100 , however , also other channels contribute : @xmath101 , which has a cross section of @xmath102 of the one for the channel through the @xmath103 production ; channels with higher resonance excitation @xcite , whose contributions are shown by long - dashed and dash - dotted lines for the isospin 1/2 and 3/2 resonances , respectively ; the nonresonant ( background ) @xmath67-wave channel @xmath104 @xcite shown by short - dashed line .
the sum of all contributions in the case of our @xmath105 cross section calculated with the set 2 and in the case of the cross section from ref .
@xcite is shown by the solid and dotted lines , respectively .
the experimental data are well described in both cases . for @xmath98 gev the set 2 is a little closer to the data , while for higher c.m .
energies the relativistic calculation of ref .
@xcite works somewhat better . for the practical applications the difference at high c.m .
energies is not important , since anyway the fritiof model is applied for the baryon - baryon collisions at @xmath106 gev in the buu program @xcite which we use in the description of pion production from heavy - ion collisions .
since our major goal is to describe the pion production from heavy - ion collisions at the beam energies @xmath107 1 a gev , we have also checked the differential cross sections @xmath108 at @xmath109 gev / c ( @xmath110 gev ) .
[ fig : sigma_vs_costh ] shows the c.m .
polar angle distribution of neutrons .
our calculation is again in a very close agreement with the relativistic treatment of ref .
@xcite and describes the data @xcite well .
we neglected the contribution of the channel @xmath111 in comparison to the data : taking into account this channel would roughly multiply our results by a factor of 10/9 , which would yield an even better agreement with the data .
the invariant mass distribution of @xmath112 pairs is shown in fig . [ fig : sigma_vs_minv ] .
we see that both models agree quite well , but the experimental distribution from @xcite is somewhat broader .
notice , that the channel @xmath111 missed in our calculations would create some broadening of the spectrum , since now the pair ( @xmath113 ) is not emitted from the same resonance .
however , the study of this effect is beyond the scope of our work .
we assume that nucleons and @xmath8 s are coupled to the scalar mean field @xmath1 by the same universal coupling constant @xmath114 @xcite .
this produces the dirac effective masses @xmath115 which are smaller than the bare ones by 300@xmath10400 mev at the normal nuclear density , as one can see from fig .
[ fig : waltst ] .
we have adopted the two parameter sets of the relativistic mean field model ( rmf ) @xcite : nl1 and nl2 , and the relativistic hartree approximation ( rha ) @xcite .
rha takes into account vacuum fluctuations which are missed in rmf .
we observe from fig .
[ fig : waltst ] , that the rmf set nl1 produces the steepest decrease of the nucleon effective mass with the baryon density , while rha gives the slowest one .
assuming , further , that the coupling constants to the vector mean field @xmath116 are also the same for nucleons and @xmath8 s @xcite , we get the kinetic four - momenta @xmath117 the effective masses and the kinetic four - momenta substitute the vacuum masses and four - momenta in the calculations .
thus , in eq .
( [ sigvac ] ) one should replace : @xmath118 the total in - medium @xmath8 width is approximated as follows : @xmath119 where @xmath120 @xmath121 is the nuclear saturation density and @xmath122 is a constant .
the first term in eq.([gammed ] ) is the @xmath8-spreading width @xcite , which describes the in - medium broadening of the @xmath8 resonance due to the absorption and rescattering processes @xmath123 , @xmath124 and @xmath125 . in ref .
@xcite the many - body calculations of the @xmath8 self - energy in nuclear matter have been performed which are in a good agreement with the empirical @xmath8-spreading potential of ref .
@xcite @xmath126 thus , the value of the constant @xmath122 consistent with refs .
@xcite is @xmath127 mev . in ref .
@xcite another value @xmath128 mev has been used in the calculation of the pion self - energy within the @xmath8-hole model . in ref .
@xcite the @xmath8-spreading width has been calculated in the relativistic meson - nucleon model including effective masses given by the walecka model .
only the @xmath123 contribution has been taken into account .
vertex corrections were neglected in ref .
@xcite , which resulted in rather big values of the @xmath8-spreading width in the calculation without effective mass ( @xmath129 mev ) .
including effective mass corrections , the @xmath8-spreading width of ref .
@xcite , first , increases with density at @xmath130 in agreement with @xmath131 mev , but than drops abruptly at higher densities due to the mean field effects .
we have checked that the variation of the parameter @xmath122 within the range @xmath132 mev leads to the changes in the in - medium cross section @xmath133 which are comparable with the ambiguity due to the choice of the effective mass parameterization ( nl1 or nl2 ) . on the other hand , we need to keep @xmath134 in order to have numerically stable results in the calculations of the @xmath8-hole lindhard function ( [ lindd1 ] ) . in the numerical results presented below we use @xmath131 mev .
the second term in eq.([gammed ] ) is the @xmath5 width taking into account the effective masses of the nucleon and @xmath8 . in this term
we have omitted the pauli blocking of the decay nucleon .
this is justified because in the real heavy - ion collision at the beam energy of 1 a gev the nuclear matter at finite temperature of about 70 mev is created , which reduces the pauli blocking effect strongly with respect to the zero temperature nuclear matter . in the c.m .
frame of colliding nucleons ( i.e. in the frame where @xmath135 ) the differential cross section in nuclear medium reads as follows ( c.f .
( [ sigvaccm ] ) ) : @xmath136 where @xmath137 .
below we will assume that the c.m .
frame of colliding particles coincides with the nuclear matter rest frame , where spatial components of the vector field disappear , thus @xmath138 .
the in - medium matrix element @xmath139 in eq .
( [ sigmedcm ] ) is given by the graphs shown in fig .
[ fig : nnnd_opem ] , where now the in - medium pion propagator @xmath140 must be put on the place of the vacuum propagator @xmath141 and the in - medium vertex corrections due to the contact interactions must be also taken into account .
the pion self - energy @xmath142 is , in general , a matrix with two isospin indices @xmath143 .
however , in this section we consider isospin - symmetric nuclear matter , where @xmath144 .
the delta function of the isospin indices will be always dropped below for brevity .
the assumption of isospin - symmetric nuclear matter results , in particular , in the validity of the vacuum relations between various isospin channels ( c.f eqs.([isorel1]),([isorel2 ] ) below ) for the in - medium cross sections . to avoid misunderstanding , we stress , however , that in the buu calculations of sect . iii we take into account the isospin assymmetry of colliding nuclei explicitly . the function @xmath145 is shown in fig .
[ fig : nnnd_pipol]a .
it includes the @xmath8-hole and nucleon - hole iterated contributions : we denoted them as @xmath146 and @xmath147 respectively .
these contributions satisfy the recurrence relations shown in fig .
[ fig : nnnd_pipol]b , c which read as : @xmath148 where the 2@xmath1492 matrix @xmath150 is @xmath151 the susceptibilities @xmath152 and @xmath153 are : @xmath154 with @xmath155 and @xmath156 being , respectively , the nucleon - hole and @xmath8-hole lindhard functions : @xmath157 here @xmath158 and @xmath159 are the nucleon and @xmath8 propagators : @xmath160 where @xmath161 are the on - shell nucleon and @xmath8 energies , @xmath162 is the nucleon occupation number .
the nucleon propagator ( [ nucprop ] ) consists of the vacuum part and the in - medium part ( @xmath163 ) .
the @xmath8 propagator ( [ delprop ] ) includes the vacuum part only , since we have neglected the presence of @xmath8 excitations in nuclear matter .
both propagators take into account the effective mass corrections . after the contour integration over @xmath164 ( c.f .
@xcite ) the nucleon - hole lindhard function ( [ lindn ] ) takes the following form : @xmath165 we have used the analytic formulas given in ref .
@xcite for the calculation of the integral ( [ lindn1 ] ) in the case of fermi distribution at zero temperature @xmath166 .
the @xmath8-hole lindhard function ( [ lindd ] ) , after integration over @xmath164 , can be expressed as @xmath167 where @xmath168 . in the calculation of the integral ( [ lindd1 ] )
we have used the expression from the appendix of ref .
@xcite ( see also the appendix in ref .
@xcite ) assuming that the invariant mass of the nucleon and pion in the argument of @xmath169 can be approximated by @xmath170 solving the system ( [ sys1 ] ) we get for the pion self - energy the following expression : @xmath171~ , \label{pipol}\end{aligned}\ ] ] where @xmath172 expression ( [ pipol ] ) coincides with the result from ref . @xcite . the renormalized @xmath75 absorption vertex operator @xmath173 is shown in fig . [ fig : nnnd_pind]a . the first graph in fig .
[ fig : nnnd_pind]a is the @xmath75 absorption vertex operator in vacuum @xmath174 ( see eq.([gama0 ] ) ) .
the second and third graphs in fig .
[ fig : nnnd_pind]a correspond to the iterated nucleon - hole and @xmath8-hole corrections which we will denote as @xmath175 and @xmath176 respectively .
the correction terms satisfy the recurrence relations shown in fig .
[ fig : nnnd_pind]b , c which can be written as : @xmath177 after solution of the system ( [ sys2 ] ) we obtain the expression for the operator @xmath173 : @xmath178 \gamma_j^{abs}(k)~. \label{gama}\end{aligned}\ ] ] in an analogous way , the renormalized @xmath76 decay vertex operator @xmath179 can be represented as a sum of the vacuum operator @xmath180 , the iterated nucleon - hole @xmath181 and @xmath8-hole @xmath182 contributions ( see fig .
[ fig : nnnd_pinn]a ) . the recurrence relations for @xmath181 and @xmath182
are shown in fig .
[ fig : nnnd_pinn]b , c .
algebraically , they look as : @xmath183 resolving ( [ sys3 ] ) one gets the renormalized vertex operator @xmath179 : @xmath184 \gamma_j^{dec}(k)~. \label{gamd}\end{aligned}\ ] ] finally , fig .
[ fig : nnnd_vcont]a shows the renormalized contact interaction @xmath185 ( we drop here the indices of external particles for brevity ) given by the sum of the bare contact interaction @xmath186 , nucleon - hole @xmath187 and @xmath8-hole @xmath188 iterated contributions .
the recurrence relations for @xmath187 and @xmath188 , shown graphically in fig . [ fig : nnnd_vcont]b , c , are : @xmath189 resolving the system ( [ sys4 ] ) we get a simple formula for @xmath190 : @xmath191 the in - medium matrix element @xmath139 is given by the sum of the direct eq.([ta ] ) and exchange eq.([tb ] ) contributions where @xmath192 and @xmath193 are replaced by the renormalized quantities .
the explicit formulas for calculation of @xmath194 are given in appendix a. fig .
[ fig : sigma_vs_ekcm_rho ] shows the total in - medium @xmath195 cross section vs the c.m . kinetic energy above the pion production threshold @xmath196 for different densities .
we compare the cross sections at fixed @xmath197 rather than at fixed c.m .
momentum , because @xmath197 is the direct measure of the collision inelasticity : e.g. , @xmath198 for all baryon densities , while the threshold value of the c.m .
momentum is density dependent .
the solid lines in the panels a , b and c of fig .
[ fig : sigma_vs_ekcm_rho ] show the results of the full calculation ( i.e. including effective masses , the pion collectivity and vertex renormalization ) for the combinations nl1-set1 , nl1-set2 and nl2-set2 , respectively .
we see that the cross section quickly drops with baryon density .
the strongest effect is obtained for the combination nl1-set2 : in this case @xmath199 drops by two orders of magnitude when the density increases from 0 up to @xmath200 .
the combination nl1-set1 produces a somewhat weaker decrease of @xmath199 with density than the combination nl1-set2 .
however , an effect of the choice of the opem parameters on the result is much smaller than the effect of the choice of the effective nucleon and @xmath8 masses .
the set nl2 gives less in - medium reduction than the set nl1 as one expects from fig .
[ fig : waltst ] .
the main reason for the in - medium reduction of the cross section is the factor of @xmath201 in front of the matrix element squared in eq .
( [ sigmedcm ] ) for the differential cross section . to demonstrate this ,
we have repeated the calculation using the nl2-set2 combination replacing this factor by its vacuum value , i.e. @xmath202 ( see dashed lines in the panel c of fig . [
fig : sigma_vs_ekcm_rho ] ) .
one observes that the cross section calculated with the vacuum factor practically does not depend on the density .
further replacements of the c.m .
momenta of the incoming and outgoing particles by their vacuum values and also of the effective by vacuum masses in the spectral function in eq .
( [ sigmedcm ] ) have no any visible effect . in order to better understand the reason for the lack of the density dependence in the calculation with the vacuum phase space factor we have also calculated the cross section using the set2 of the opem parameters with vacuum masses both in the matrix element and phase space factors ( panels e and f of fig .
[ fig : sigma_vs_ekcm_rho ] ) .
panels e and f of fig .
[ fig : sigma_vs_ekcm_rho ] show the results with and without pion collectivity and vertex renormalization effects , respectively .
namely , in the calculation presented in panel f we have put all the lindhard functions equal to zero .
however , the @xmath8-spreading width is still present in the results shown in panels e and f. ( i ) as one can see from panel e , the cross section drops with density even without any inclusion of the effective masses . by comparing the solid lines in panel e with the dashed lines in panel c
we conclude that the inclusion of the effective masses in the matrix element increases the cross section .
( ii ) from panel f we observe that the @xmath8-spreading width in the spectral function reduces the cross section slightly . as discussed above , this effect is comparable with the ambiguity given by the choice of the effective mass parameterization .
( iii ) by comparing the curves in panels e and f one sees that the pion collectivity and vertex renormalization reduce the cross section rather noticeably .
therefore , the density independence observed for the calculation with the vacuum phase space factor only ( dashed lines in panel c ) is , basically , due to a counterbalance of the effects ( i ) and ( iii ) .
panel d of fig .
[ fig : sigma_vs_ekcm_rho ] shows the nl2-set2 calculation with the lindhard functions putted to zero in the matrix element ( solid lines ) and also the calculation using the matrix element of ref .
@xcite with the in - medium nucleon and @xmath8 masses from the nl2 set ( dashed lines ) . as expected , both calculations are in a very close agreement since the nucleon - hole and @xmath8-hole lindhard functions are absent in both cases .
a comment is in order here on the relation to the results of ref .
@xcite where it was shown neglecting effective mass corrections that the in - medium cross section @xmath199 grows with baryon density .
we have repeated the calculations of @xcite with the parameter set b ( @xmath203 gev , @xmath204 ) for the proton beam energy @xmath205 gev . following ref .
@xcite we have kept only the @xmath8-hole lindhard function in the quasistatic approximation with zero width of the @xmath8 resonance @xmath206 where @xmath207 . in the @xmath8 spectral function the @xmath8 width has been also put to zero @xcite . within these approximations
we have reproduced the result of ref .
@xcite on the in - medium enhancement of @xmath199 ( see crosses in fig . [
fig : sigma_vs_rho ] ) .
the inclusion of the in - medium @xmath8 width reduces the in - medium enhancement ( see filled boxes in fig . [
fig : sigma_vs_rho ] ) , which has been also demonstrated earlier in ref .
we have found that taking into account the nucleon - hole lindhard function makes @xmath199 a decreasing function of the density at @xmath208 in the calculations with vacuum masses ( filled circles in fig .
[ fig : sigma_vs_rho ] ) . in the calculations with effective masses
the nucleon - hole lindhard function has a very small effect . for the choice of the landau - migdal parameters and of the cut - off factor from set2 the cross section decreases faster with density ( triangles in fig .
[ fig : sigma_vs_rho ] ) . introducing the effective masses
amplifies the in - medium reduction effect strongly ( see rombuses in fig . [
fig : sigma_vs_rho ] ) . summarizing all these findings
, we can state that by just changing the factor in front of the matrix element in ( [ sigvaccm ] ) to its in - medium value takes care of most of the in - medium corrections on the @xmath8-production cross section .
this finding offers a way that can easily be implemented into any numerical realization . in fig .
[ fig : dsigdcos ] we show the in - medium differential cross section of the outgoing neutron at various densities for @xmath209 gev which corresponds to @xmath210 gev / c at @xmath211 . for the calculation without effective mass ( lower right panel ) the shape of the angular distribution does not depend , practically , on the density for @xmath212
: only changes for forward / backward angles are visible . however , in the calculations with effective masses ( all other pannels ) the cross section becomes more isotropic with increasing density in qualitative agreement with the dirac - brueckner calculations of the elastic nn cross section in nuclear matter @xcite .
the in - medium modified @xmath9 cross section has been implemented in the buu program of the version described in ref . @xcite .
more details can be found in ref .
the cross sections for another isospin channels are related to the @xmath9 cross section by the clebsch - gordan coefficients which gives : @xmath213 besides the @xmath8-excitation , there are another channels of the pion production in a nn collision . at collision energies above the pion
production threshold @xmath214 not only the @xmath4 , but any other resonance @xmath215 can be excited : @xmath216 where @xmath215 stands for n@xmath217(1440 ) , n@xmath217(1535 ) , n@xmath217(1650 ) ... ( see ref .
@xcite for the full list of resonances implemented in the buu code ) . at @xmath218
the double @xmath8 production channel opens : @xmath219 in addition , at collision energies slightly above the pion production threshold the @xmath67-wave direct pion production mechanism is important : @xmath220 in analogy with the density modification of the @xmath8 resonance cross section , we have also modified the processes ( [ nnnr]),([nndd ] ) and ( [ nnnnpi ] ) keeping , however , the constant density independent matrix elements as determined in ref .
@xcite . for the in - medium cross section for the process ( [ nnnr ] )
we apply eq .
( [ sigmedcm ] ) , where any higher resonance replaces the @xmath8 resonance and @xmath194 is now a constant depending on the isospin channel .
the effective mass of a higher resonance @xmath215 is supposed to be @xmath221 i.e. we assume the same mass shift for all baryon resonances .
the spectral function of a higher resonance @xmath215 is given by eq .
( [ spfunmed ] ) with replacement `` @xmath8 ''
@xmath222 `` @xmath215 '' , where the medium - modified total width is a sum of partial widths : @xmath223 where @xmath224 .
@xmath225 is the pole mass of a resonance @xmath215 and @xmath226 here , the lower indices `` @xmath227 '' and `` @xmath55 '' mean the outgoing baryon and meson for the @xmath228-th decay channel : ( @xmath229 ) = ( @xmath230 ) , ( @xmath231 ) , ( @xmath232 ) , ( @xmath233 ) , ( @xmath234 ) , ( @xmath235 ) , ( @xmath236 ) , ( @xmath237 ) , ( @xmath238 ) . @xmath239 and @xmath240 are the spectral functions of the outgoing particles , @xmath241 is a blatt - weisskopf barrier penetration factor corresponding to the relative angular momentum @xmath242 of the outgoing baryon and meson .
( [ gamrst]),([gamrist]),([rhoi ] ) are the in - medium modification of the vacuum widths from ref .
we have to remark only , that also for the outgoing @xmath81 we take in eq .
( [ rhoi ] ) an in - medium spectral function @xmath243 .
this is nessary in order to keep the decay channel @xmath244 open in nuclear matter , since the resonance @xmath215 has now reduced mass .
however , we did not modify any cross sections involving strange particles which is out of scope of our work . for the in - medium cross section of the process @xmath245
we apply the following formula : @xmath246 where the matrix element @xmath247 depends on the isospins of incoming and outgoing particles only .
@xmath248 if the i - th and j - th particles are in the different ( same ) isospin states . in the vacuum
the cross section ( [ sigmedddcm ] ) is the same as in ref .
@xcite .
finally , we have also introduced effective mass modifications for the process of the direct @xmath67-wave pion production @xmath249 which can be written in the form of the dalitz plot @xcite : @xmath250 where @xmath251 and @xmath252 are the invariant masses squared of the pairs of outgoing particles @xmath253 and @xmath254 respectively .
@xmath255 is the vacuum matrix element squared which depends on the vacuum c.m .
energy @xmath256 ( see eq.([srtst ] ) below ) and on the isospins of the involved particles . from eq .
( [ sigmedswave ] ) one gets the following relation between the in - medium and vacuum cross sections : @xmath257 where @xmath258 is the vacuum cross section parameterized in ref . @xcite . in order to maintain detailed balance , one must modify also the inverse processes @xmath259 , @xmath260 , @xmath261 , @xmath262 which are obviously essential for pion reabsorption
simultaneously with the direct reactions .
we collect the formulas for the in - medium cross sections of the inverse processes in appendix b. since the mean field included in the buu model is a skyrme - like which is different from the rmf model , a recipe must be found how to match the buu kinematics of a nucleon - nucleon collision with the kinematics employing dirac effective masses .
one way is to take the incoming particle momenta from buu and construct @xmath263 by replacing the bare masses by the effective ones .
however , as we discussed in sect .
ii , if the c.m . momentum is fixed , a nn collision which is subthreshold for pion production in the vacuum may get above threshold in the nuclear medium .
this would create practical difficulties in ensuring energy conservation for the outgoing particles .
therefore , we have used the c.m .
kinetic energy above the pion production threshold from buu as an input to the calculation of the in - medium cross sections .
namely , for each nn collision we find @xmath264 , where @xmath256 is calculated with the incoming nucleon momenta from buu using the vacuum dispersion relation for nucleon energies .
the in - medium c.m .
energy @xmath265 is determined by requiring that @xmath266 , i.e. @xmath267 correspondingly , the c.m .
momentum of the incoming particles gets changed : @xmath268 . replacing only the vacuum cross sections by the in - medium ones in buu
would , however , not correspond to a consistent description of the resonance production / absorption rates because @xmath269 where @xmath270 is the transition probability per unit time .
@xmath271 is the relative velocity and @xmath272 are the c.m .
energies of colliding particles 1 and 2 . due to the difference between the mean fields in buu and in the rmf model ,
the relative velocities of colliding particles are , generally , different in both models too .
therefore , in order to obtain the rates corresponding to the rmf model , we have multiplied the cross sections calculated above by the factor @xmath273 , where @xmath274 is the relative velocity of the colliding particles as given by buu .
the buu results have been obtained by applying the soft momentum dependent mean field ( sm , @xmath275 mev , see refs .
@xcite for details ) in the hamiltonian propagation of baryons .
[ fig : pidelta ] shows the time dependence of the sum of the pion and @xmath8 multiplicities in a central collision of au+au at 1.06 a gev . at @xmath276 fm / c practically all the @xmath8 s
have already decayed , so that the final pion multiplicity can be determined from fig .
[ fig : pidelta ] .
a standard calculation without any in - medium effects in the cross sections ends up with 68 pions , which is much more than the experimental value of @xmath277 @xcite .
using the medium modified cross sections reduces the pion multiplicity .
a bigger reduction corresponds to a steeper dropping effective mass with the density ( c.f .
[ fig : waltst ] ) .
thus , the smallest pion multiplicity of @xmath278 is obtained within the nl1-set2 combination .
it is interesting that the combinations nl1-set1 and nl2-set2 produce nearly equal pion numbers : 45 and 43 , respectively .
this is explained by the fact , that the maximum central density reached in central au+au collisions at 1 a gev is about @xmath279 , while the cross sections given by the nl1-set1 and nl2-set2 combinations are similar at @xmath280 ( see fig . [
fig : sigma_vs_ekcm_rho ] ) . in all calculations below we will always use the set2 of the opem parameters , thus , we will drop for brevity the label `` set2 '' . fig . [
fig : pimul ] shows the average total pion multiplicity @xmath281 divided by the average participant number @xmath282 as a function of the total mass number @xmath283 of colliding nuclei in comparison to the data @xcite .
the systems ca+ca , ru+ru and au+au were studied at the beam energies of 400 , 1000 and 1500 a mev .
following @xcite , we have fitted the total pion multiplicity as a function of the participant number @xmath284 by a straight line in the range @xmath285 .
the participant number @xmath284 has been extracted from the geometrical overlap of colliding nuclei for a given impact parameter taking into account smooth density profiles as described in @xcite .
first , we notice that there is no a single set of parameters which is able to describe all data points in fig .
[ fig : pimul ] simultaneously .
the standard calculation agrees with the data at 400 a mev , but it clearly overpredicts the experiment at 1000 a mev and , for heavy systems , also at 1500 a mev . the nl1 parameterization works for the au+au system at 1000 and 1500 a mev , however , it is below the data at 400 a mev .
furthermore , nl1 is unable to describe the experimental slopes at 1000 and 1500 a mev .
rha is in the best agreement with data at 400 a mev .
surprisingly , for the au+au system at 400 a mev rha gives more pions than the standard calculation .
this , probably , can be explained by the reduced reabsorption of @xmath8 s in rha due to the in - medium reduced cross section @xmath286 .
rha is also rather close to the experiment at 1000 a mev , while at 1500 a mev the system mass dependence is too flat within rha .
the best overall description of pion multiplicities seems to be reached within the nl2 parameterization .
more precise data , in particular , at the lowest energy are needed for definite conclusions , however .
[ fig : pirat ] shows the ratio of the average @xmath287 and @xmath288 multiplicities vs the total mass number of a system for ca+ca , ru+ru and au+au at 400 , 1000 and 1500 a mev in comparison to the data @xcite .
the calculated ratio @xmath289 practically does not depend on the in - medium modifications of the cross sections , since these modifications do not change the ratios between the cross sections for various isospin channels . for ca+ca and ru+ru all the calculations agree with the data , while for the au+au system
there is a systematic tendency to produce about 10 % too many @xmath287 s in our calculations .
probably , the @xmath289 ratio would be described better if to take into account the isospin dependence of the in - medium part of the nucleon propagator ( [ nucprop ] ) .
[ fig : piptsp_au106au ] shows the inclusive transverse momentum spectra of pions at midrapidity for the au+au system at 1.06 a gev . the standard calculation ( dotted line )
overpredicts the data at @xmath290 mev / c .
nl1 ( dashed line ) leads to a good description at @xmath291 mev / c , but it underpredicts the experiment at low @xmath292 .
almost the same result has been obtained earlier in @xcite using the phenomenological density dependent quenching factor for the @xmath293 cross sections .
a better overall description is reached within the nl2 parameterization ( dash - dotted line ) , however the yields of @xmath287 s and @xmath294 s are still underpredicted in the region of small transverse momenta .
this is improved by taking into account the pion off - shellness in buu @xcite ( solid line ) , which enhances the soft pion yield keeping the total pion number practically unchanged . in order to see how the density modified cross sections work in the case of light colliding nuclei we selected the taps data on @xmath294 production from collisions of c+c at 0.8 , 1.0 and 2.0 a gev @xcite .
[ fig : pinum_cc ] shows the impact parameter averaged @xmath294 multiplicity @xmath295 in a narrow rapidity interval near @xmath296 .
standard calculation is in the best agreement with the data for @xmath295 , while the nl1 and nl2 calculations are below the data for the beam energies of 0.8 and 1 a gev .
finally , fig .
[ fig : dmtpi_cc_new ] shows transverse mass spectra of @xmath294 s at midrapidity for the c+c collisions at 0.8 , 1 and 2 a gev .
the standard calculation provides a quite good description of the data @xcite .
in - medium modifications of the cross sections do not destroy a good description at 2 a gev and also improve agreement with experiment for large transverse masses for 0.8 and 1 a gev , but tend to underpredict the data at small @xmath297 for 0.8 and 1 a gev .
in this work we have calculated the in - medium @xmath298 cross section within the opem taking into account the exchange pion collectivity , vertex corrections by the contact nuclear interactions and the effective masses of the nucleon and @xmath8 resonance .
\(i ) we have observed that even without the effective mass modifications the cross section decreases with the nuclear matter density at high densities if one takes into account the in - medium @xmath8 width and includes the @xmath299 lindhard function in the calculations ( fig .
[ fig : sigma_vs_rho ] ) .
\(ii ) inclusion of the effective mass modifications for the nucleons and @xmath8 s leads to an additional strong reduction of the cross section .
\(iii ) the fopi data @xcite on the total pion multiplicity from the systems ca+ca , ru+ru and au+au at 0.400 , 1.000 and 1.500 a gev seem to require a dropping effective mass with the baryon density ( nl2 version of the rmf model ) in combination with the universal value of 0.6 for all the landau - migdal parameters in the spin - isospin channel .
we note that this still depends somewhat on the size of the @xmath8-spreading width used ( 80 mev @xmath300 @xcite ) .
more precise data on the total pion multiplicity are needed in order to better select the model for the density dependence of the dirac effective masses of the baryons .
the effect of the medium modifications of the @xmath293 cross sections on the pion multiplicity depends also on the assumption about other channels of the pion production / absorption in nn collisions , most importantly , on the @xmath67-wave direct channel @xmath301 . including the effective mass modifications in the @xmath302 channel _
only _ , does not lead to strong enough reduction of the observed pion multiplicity , since then more pions are produced in the @xmath67-wave channel .
the in - medium modifications of the higher resonance cross sections do not influence the pion production at 1 - 2 a gev collision energy sensitively : other particles like @xmath303 and @xmath304 mesons are , probably , more sensitive to higher resonance in - medium modifications .
the calculations with the in - medium modified cross sections slightly underpredict the low transverse mass @xmath294 yield at midrapidity for c+c collisions at 0.8 and 1.0 a gev ( fig .
[ fig : dmtpi_cc_new ] ) .
this implies , that with decreasing mass number of the colliding system the in - medium corrections disappear faster than in our model .
the reason could be a strong nonequilibrium momentum distribution at the beginning of collision : instead of one fermi sphere there are two nonoverlapping fermi spheres , which is not taken into account in our calculation of the effective mass .
indeed , the nonequilibrium momentum distribution reduces the scalar density , and , therefore , increases the effective mass .
this effect is expected to be stronger for lighter systems , since smaller baryon density is reached in this case .
another open problem is the general tendency of the transport models to overpredict pion multiplicity in heavy colliding systems at ags energies ( c.f .
this problem can not be solved by just density modifications of the resonance production / absorption cross sections , since most pions are produced by the string mechanism here . in - medium modifications of the fritiof model , thus , could help to solve this problem . numerous stimulating discussions with m. post and dr .
l. alvarez - ruso are gratefully acknowledged .
the authors are grateful to markus post for the reading of the manuscript before publication , critical comments and suggestions and to luis alvarez - ruso for pointing out ref .
@xcite which has influenced our results .
for the calculation of @xmath194 it is convenient to introduce the renormalized coupling constants @xmath305
f_\delta(k^2 ) ~ , \label{fdren } \\ \tilde{f}(k ) & \equiv & { \det}^{-1}(\hat a(k ) ) [ 1 + ( g_{\delta\delta}^\prime - g_{n\delta}^\prime)\chi_\delta(k ) ] f(k^2 ) \label{fren}\end{aligned}\ ] ] and the renormalized landau - migdal parameter @xmath306 [ 1 + ( g_{\delta\delta}^\prime - g_{n\delta}^\prime)\chi_\delta(k)]}~. \label{gren}\ ] ] in terms of @xmath307 , @xmath308 and @xmath309 the renormalized vertex operators and the renormalized contact interaction have the same form as the vacuum ones : @xmath310 [ \chi^\dag(3 ) \sigma_\alpha \mbox{\boldmath${\mathbf\tau}$\unboldmath } \chi(1)]~. \label{vcont1}\end{aligned}\ ] ] notice , that @xmath307 , @xmath308 and @xmath309 are complex - valued quantities .
substituting ( [ gama1]),([gamd1 ] ) and ( [ vcont1 ] ) into ( [ ta ] ) on the place of corresponding vacuum quantities and also replacing @xmath88 ( see eq.([k4 ] ) ) in eq.([gama1 ] ) one can get after some algebra : @xmath311
\left(-{t \over { \bf k}^2}\right)~ , \label{tamedsq}\end{aligned}\ ] ] where the term @xmath312 is introduced in order to restore the lorentz invariance , as discussed in sect .
the exchange term @xmath313 is given by eq.([tamedsq ] ) with replacements @xmath314 , @xmath86 and @xmath315 ( see eq.([k4p ] ) ) .
for the spin - averaged interference term we get : @xmath316 = - { \tilde{f}_\delta^*(k^\prime)\tilde{f}^*(k^\prime ) \tilde{f}_\delta(k)\tilde{f}(k ) \over m_\pi^4 } \nonumber \\ & \times & \left\ { { 1 \over 3 } [ 2({\bf k_4^\prime \cdot k_4})({\bf k^\prime \cdot k } ) + ( { \bf k_4 \cdot k } ) ( { \bf k_4^\prime \cdot k^\prime } ) - ( { \bf k_4 \cdot k^\prime})({\bf k_4^\prime \cdot k } ) ] d^*(k^\prime ) d(k ) \right . \nonumber \\ & + & \left .
{ 4 \over 3 } ( { \bf k_4^\prime \cdot k^\prime } ) d^*(k^\prime ) \tilde{g}(k ) + { 4 \over 3 } ( { \bf k_4 \cdot k } ) d(k ) \tilde{g}^*(k^\prime ) + 4\tilde{g}^*(k^\prime)\tilde{g}(k ) \right\ } { ( tu)^{1/2 } \over ( { \bf k}^2{\bf k^\prime}^2)^{1/2 } } + c.c.~. \label{tintmedsq}\end{aligned}\ ] ]
we use below the detailed balance principle , i.e. the equality of the matrix elements of the direct and inverse processes , in order to get the cross sections for the reactions @xmath259 , @xmath260 , @xmath261 and the pion @xmath67-wave absorption rate by two - nucleon pairs @xmath262 .
the in - medium differential cross section for the process @xmath259 is given by the expression ( c.f .
eq.([sigmedcm ] ) ) : @xmath317 where @xmath194 is calculated in appendix a ( eqs . ( [ tamedsq]),([tintmedsq ] ) ) .
@xmath318 is the spin of the @xmath8-resonance . for the process @xmath260 , where @xmath215 means one of the higher resonances
, one should simply replace in eq.([dnnn ] ) `` @xmath8 '' by `` @xmath215 '' : now @xmath319 and @xmath320 are the effective mass and spin of the resonance @xmath215 , respectively .
@xmath194 is now a constant depending only on the isospin channel .
the pion @xmath67-wave absorption rate by the two - nucleon pairs is : @xmath322 where @xmath323 is the nucleon spin degeneracy factor , @xmath324 are the nucleon occupation numbers .
the sum in eq.([swpimed ] ) is taken over the isospin projections of the incoming nucleons 1 and 2 .
the vacuum matrix element squared @xmath325 depends on the vacuum c.m .
energy and on the particle isospins .
the integral ( [ swpimed ] ) is performed by the monte - carlo method as described in ref .
@xcite .
s. teis , w. cassing , m. effenberger , a. hombach , u. mosel and gy .
wolf , z. phys .
a * 356 * , 421 ( 1997 ) ; + s. teis , w. cassing , m. effenberger , a. hombach , u. mosel and gy .
wolf , z. phys .
a * 359 * , 297 ( 1997 ) .
upper panel : the total cross section @xmath105 in vacuum as a function of the c.m .
energy @xmath256 .
calculations are performed using the set 1 ( dashed line ) and set 2 ( solid line ) of the opem parameters ( see table 1 ) .
results obtained within the model of ref .
@xcite are shown by dotted line .
lower panel : vacuum cross section of a single @xmath288 production in @xmath327 collisions vs @xmath256 .
results with the @xmath105 cross section calculated with the set 2 and using the model of ref .
@xcite are shown by solid and dotted lines respectively . the @xmath67-wave background , isospin 1/2 and 3/2 ( excluding @xmath328 ) resonance contributions @xcite are shown by short - dashed , long - dashed and dash - dotted lines respectively .
solid and dotted lines show the incoherent sum of the @xmath105 cross section multiplied by 10/9 and the background plus isospin 1/2 and 3/2 higher resonance contributions .
the data are from ref . @xcite . ]
outgoing neutron c.m .
polar angle distribution in the reaction @xmath104 at @xmath109 gev / c . in the calculations only the @xmath105 channel is taken into account .
the curves are labeled as in the upper panel of fig .
[ fig : sigma_vs_plab ] .
the histogram shows the data from ref .
@xcite . ]
invariant mass distribution of outgoing @xmath288 and proton for @xmath327 collisions at @xmath109 gev / c . in the calculations
only the @xmath105 , @xmath329 reaction channel is taken into account . the data from ref .
@xcite are shown by the histogram . ]
ratio of the effective nucleon mass to the vacuum nucleon mass vs the ratio of the baryon density to the nuclear saturation density .
dashed , solid and dash - dotted lines show the calculations using nl1 , nl2 and rha respectively . ]
( a ) pion self - energy @xmath330 .
( b),(c ) the recurrence relations for the @xmath331 and @xmath332 iterated contributions ( see eqs.([sys1 ] ) ) .
arrows are not shown on the @xmath331 loops , since each of these loops is the sum @xmath67- and @xmath333-channel graphs ( with anticlockwise and clockwise directed arrows , respectively ) . ] in - medium @xmath105 cross section as a function of the kinetic energy above the pion production threshold @xmath197 ( see eq.([qstar ] ) ) for various parameter sets ( see text for details ) . in all cases , excepting the dashed lines in panel ( c ) , lines from the uppermost to the lowermost correspond to the baryon densities @xmath334 and @xmath335 , respectively .
for the dashed lines in panel ( c ) the density order is opposite . ]
density dependence of the ratio of the in - medium and vacuum total cross sections @xmath105 for @xmath336 gev ( in vacuum this corresponds to the proton beam energy of 0.8 gev ) .
line marked with crosses corresponds to the approximations of ref .
@xcite with the parameter set b @xcite .
boxes and circles show the calculations including the in - medium @xmath8 width without and with the @xmath299 lindhard function contribution respectively .
triangles and rombuses show the results for the set2 of the opem parameters without and with inclusion of the effective mass , respectively . ]
neutron c.m .
polar angle dependence of the in - medium differential @xmath105 cross section .
the lines from the uppermost to the lowermost correspond to the baryon density @xmath334 and @xmath335 , respectively .
calculations are performed for @xmath209 gev . ]
time dependence of the sum of the pion and @xmath8 multiplicities in the central collision of au+au at 1.06 agev .
the standard buu calculation is shown by the dotted line .
the dash - dotted , solid , short - dashed and long - dashed lines show the results with the in - medium modified resonance production / absorption cross sections by using the combinations rha - set2 , nl2-set2 , nl1-set1 and nl1-set2 , respectively . ] total pion yield per participating nucleon for different energies and colliding systems .
dotted , dash - dotted , solid and dashed lines correspond to the standard , rha , nl2 and nl1 calculations , respectively .
the data points are from ref .
the results of the data analysis a ( b ) are represented by filled circles ( stars ) @xcite . ]
pion inclusive transverse momentum spectra at midrapidity for the reaction au+au at 1.06 agev .
dotted lines show the standard buu calculation . dashed and dash - dotted lines show results with in - medium modified resonance production / absorption cross sections applying the nl1 and nl2 parameterizations , respectively .
solid lines correspond to the nl2 parameterization taking into account pion off - shellness .
the normalized rapidity intervals in which spectra are extracted are @xmath337 for @xmath338 and @xmath339 for @xmath294 , where @xmath340 .
the data are from ref . @xcite . ]
the average @xmath294 multiplicity in a narrow rapidity interval near midrapidity as a function of the beam energy for c+c collisions .
dotted , dashed and solid lines correspond to the standard , nl1 and nl2 calculations , respectively .
the rapidity intervals are @xmath341 at @xmath205 and @xmath342 a gev , and @xmath343 at @xmath344 a gev .
the data are from ref . @xcite . ]
transverse mass spectra of @xmath294 mesons in the narrow rapidity intervals near midrapidity ( see caption to fig .
[ fig : pinum_cc ] ) for the @xmath345c+@xmath345c system at 0.8 , 1.0 and 2.0 a gev .
curves are denoted as in fig .
[ fig : piptsp_au106au ] .
the data are from ref . @xcite | we present calculations of the @xmath0 cross section in nuclear matter within the one - pion exchange model taking into account pion collectivity , vertex renormalization by the contact nuclear interactions and dirac effective masses of the baryons due to coupling with the scalar @xmath1 field . introducing the dirac effective masses leads to an in - medium reduction of the cross section .
the experimental data on pion multiplicities from the collisions of ca+ca , ru+ru and au+au at @xmath2 a gev are well described by buu calculations with the in - medium cross section . |
high energy astrophysical sources show a temporal variability in a broad spectral range . the x - ray ( 1 - 10 kev ) and hard x - ray ranges ( 10 - 100 kev ) are especially important observing windows to understand the temporal characteristics of high energy sources which involve spectral changes .
for instance , a black hole candidate shows state changes in its flux and spectrum ; in the quiescent state the x - ray emission is dominated by the hard emission , whereas in outburst the soft emission from the accretion disk becomes dominate .
some low mass x - ray binaries produces bright bursts in x - rays , so called x - ray bursts , which are caused by thermonuclear flashes of accreting material on the surface of the neutron star .
an accreting x - ray binary pulsar sometimes shows intense outbursts in x - rays when the neutron star passes the disk or dense stellar wind of its companion .
cyclotron resonance lines , one of the direct observational approaches to measure the magnetic field of a neutron star , are most easily observable in the hard x - ray range from binary pulsar during outburst .
an isolated neutron star with a high magnetic field , a so called magnetar , occasionally shows an outburst with multiple short duration bursts in hard x - rays believed originate from a large release of its internal magnetic energy . the time - domain astronomy which has been revolutionized by _
@xcite has become a frontier field of astronomy .
the burst alert telescope ( bat ; @xcite ) onboard _ swift _ has been monitoring the hard x - ray sky ( 14 - 200 kev ) thanks to its wide field of view since 2004 . on the other hand , the gas slit camera ( gsc ; @xcite ) on the _ maxi _ mission @xcite has been observing the sky in softer energy band ( 2 - 30 kev ) since 2009 .
both instrument teams are providing the light curve data in real - time to the public @xcite .
however , the swift / bat transient monitor @xcite is limited to a single energy band of 15 - 50 kev because the pipeline is using the data extracted in a single 15 - 50 kev band on - board ( the data product called the bat scaled - map data ) .
furthermore , there is no realtime transient monitor producing light curves in the full dynamic range of the bat and the gsc data with the same time axis . by combining the data of the bat and the gsc , we are able to construct broadband light curves of high energy transient sources with a high scientific merit .
for example , the spectral state changes of several black hole candidates are investigated intensively using the bat and the gsc data ( xte j1752@xmath0223 , e.g. , @xcite ; swift j1753.5@xmath00127 , e.g. , @xcite ; swift j1910.2@xmath00546 , e.g. , @xcite ; gx 339@xmath04 , e.g. , @xcite ) .
the state transition of bright low - mass x - ray binaries have also been studied in detail using both bat and gsc data ( e.g. , @xcite . ) . using both the _ maxi _ and the _ integral _ data , there is the web page for monitoring x - ray and hard x - ray activities of high - mass x - ray binaries .
therefore there is a great demand to gather broad band light curves of high - energy transient sources from multiple missions and present them in a consistent format along a single time axis . in this paper
, we introduce the broadband transient monitor utilizing the bat and gsc data .
this transient monitor covers the dynamic range from 2 kev to 200 kev , which is ideal to monitor high energy transients in a broad spectral coverage .
the paper is organized as follows .
we present the analysis method in 2 . in
3 , our broadband transient monitor is introduced .
we further discuss our broadband transient monitor in 4 .
the flowchart of our pipeline is shown in figure [ fig : flowchart ] .
as the initial step to construct the broadband transient monitor , independent pipelines were developed just to mirror the _
swift _ bat data to our server .
this step is rather crucial for the entire process because it takes a significant amount of time to download the bat data from the _ swift _ data archive center in the u.s .
( swift data center ( sdc ) and heasarc ) to japan .
since the _ swift _ data are initially stored at the sdc , and then , archived to the heasarc seven days after the observation , two separate pipelines were developed to mirror both sdc and heasarc data , which are available to the public from our web server ( the sdc data mirror and the heasarc data mirror ) located in japan .
note that only the swift / bat related data are mirrored from the heasarc to our server .
the script to mirror the sdc data is running once an hour , whereas the script to mirror the heasarc archive runs once a day .
the basic bat data analysis is performed using the heasoft software package .
the batsurvey script is used to process the bat survey ( detector plane histogram ( dph ) ) data .
the default eight energy bands ( 14 - 20 kev , 20 - 24 kev , 24 - 35 kev , 35 - 50 kev , 50 - 75 kev , 75 - 100 kev , 100 - 150 kev and 150 - 195 kev ) and the original time resolution of the dph data ( timesep = `` dph '' ) were specified in the script .
the typical exposure time of the original dph data is five minutes .
products are made for 146 sources which were flagged as bright hard x - ray sources in the bat 70 month all - sky hard x - ray survey @xcite .
the data after august 2009 , when _ maxi _ and _ swift _ operations overlap , were processed .
the batsurvey script produces so - called ` level 2 ' catalogs for each source , which contain the extracted count - rate and the incident angle of the source .
once the batsurvey process is completed , all the level 2 catalogs ( final outputs of the batsurvey script ) are merged using batsurvey - catnum script and time sorted by ftsort for all 146 sources .
the bat count rates extracted by the batsurvey script must be corrected for the energy dependent vignetting of incoming photons @xcite . to model this off - axis effect in the count rate at each energy band , we processed the survey data between 2004 and 2005 , when the crab nebula was in the field of view at various incident angles .
figure [ fig : crab_offaxis ] illustrates the significant energy dependent systematic effect between the crab count rate and incident angle in the 15 - 20 kev and the 100 - 150 kev band , as an example .
a quadratic function was used to fit the trend between the count rate and the incident angle for all eight energy bands ( table [ tbl : offaxis_corr ] ) . in each band , the count rate is corrected by the rate of the on - axis crab rate to the estimated crab rate at the given incident angle .
next , the original eight energy bands are binned to four energy bands ( 14 - 24 kev , 24 - 50 kev , 50 - 100 kev and 100 - 195 kev ) . and
then , a one day averaged and a 90-minutes ( 1 orbit ) averaged light curves are generated using rebingausslc .
the _ maxi _ gsc one day and 90 minutes light curves are downloaded from the _ maxi _ public web page for the common sources of the bat 146 bright sources in the input catalog of the bat data process and the 369 ( as of july 2015 ) sources listed in the _ maxi _ public web page . the number of common sources in the current monitor is 106 .
this limitation mainly comes from the number of sources in the bat input catalog .
however , since all the created bat sky images are stored in our computer , only the source extraction tool , batcelldetect , is needed to run through all the archival images to add a new source from the bat data .
we have a plan to add the sources which were detected in outburst in the bat transient monitor @xcite in the past six years to our transient monitor pipeline .
the light curves of the bat and the gsc are combined and plots are generated using the python matplotlib module .
the bat light curve data are available in fits format .
the interactive light curve based on the python mpld3 module is also available , so that a user can move and zoom - in the light curve interactively .
our broadband transient monitor is available to the public from the web server at aoyama gakuin university : http://yoshidalab.mydns.jp / bat_gsc_trans_mon/. currently , the broadband light curves of 106 known sources are accessible from the web page .
the web page updates @xmath13 times a day depending on the amount of the data which need to be processed .
the products of our bat - gsc broadband transient monitor are summarized in table [ tbl : bb_product_summary ] . here
, we highlight the products of several sources in our broadband transient monitor .
figure [ fig : bb_lc_cygx1 ] and [ fig : bb_lc_grs1915 ] show our broadband light curves of the black hole binaries cygnus x-1 and grs 1915 + 105 .
those light curves clearly show multiple spectral state transitions between low - hard " and high - soft " states , which are believed to indicate a change in the geometry of the accretion disk ( e.g. , @xcite ) .
for example , the last clear state change was happening around mjd 57125 for cygnus x-1 and around mdj 56252 for grs 1915 + 105 .
as can be seen in the light curves , the borderline energy which the emission becomes brighter or dimmer when the source is in the high - soft state is around 10 kev for cygnus x-1 and 15 kev for grb 1915 + 105 .
figure [ fig : bb_lc_gx339 ] shows the long - term temporal behavior of gx 339 - 4 , a galactic transit black hole with a low - mass companion . during the giant outburst in the year 2010 ( e.g. , mjd 55200 - 55600 ) , the burst emission showed a strong hard - to - soft evolution .
the initial hard emission was visible up to the highest energy band of 100 - 195 kev .
this hard x - ray emission episode is dominated by comptonized photons from the accretion disk ( e.g. , @xcite ) .
the @xmath2-ray blazar , mrk 421 , showed several flares visible up to 100 kev ( figure [ fig : bb_lc_mrk421 ] ) .
since the bat energy range is located at the dip between the two broad peaks in the spectral energy distribution of mrk 421 ( e.g. , @xcite ) , the bat hard x - ray emission is not that evident compared to the soft x - ray band of the gsc .
however , during the outburst , the hard x - ray emission is clearly visible in the bat data , and the broad - spectral properties of the source can be investigated using our monitor .
figure [ fig : bb_lc_ngc4151 ] shows the light curve of the seyfert 1.5 galaxy ngc 4151 which has a hard continuum in its spectrum @xcite . unlike previous examples , its emission and the temporal variability are clearly visible in the bat data rather than the gsc data . in figure [
fig : bb_lc_a0535 ] , the several giant outburst episodes are visible in the broadband light curve of a high - mass x - ray pulsar a0535 + 262 .
since the emission during the giant outbursts were so intense , the temporal profile up to 100 kev is clearly visible even in the one orbit light curve ( right panel of figure [ fig : bb_lc_a0535 ] ) .
a recent study showed that the energy of the cyclotron resonance line of a0535 + 262 , which was found typically around 45 kev , increased when the flux was high @xcite .
thus , it is important to monitor those cyclotron sources in a broad energy band to understand the transient spectral features in various flux levels .
as can be seen in those examples , our transient monitor includes various types of sources and can monitor interesting temporal and spectral stages of high energy sources .
we constructed the broadband transient monitor using the _ swift _ bat and the _ maxi _ gsc data and made it available to the public .
this work is the first attempt to analyse the bat survey data in real - time for constructing multi - band light curves .
although the updating frequency of the light curve is still not ideal , we can now run the pipeline several times a day in real - time . although it varies based on the _ swift _ pointing , a typical exposure time of an individual object for the bat is roughly four hours per day ( about eight hours at the highest day ) . applying the bat survey sensitivity equation described in @xcite
, this typical exposure corresponds to a 5-sigma sensitivity of @xmath110 mcrab ( 14 - 195 kev ) .
on the other hand , the daily averaged 5-sigma sensitivity of the gsc is @xmath115 mcrab in the 4 - 10 kev band @xcite .
therefore , the daily sensitivity of the bat and the gsc are comparable .
therefore , it is a good match to combine the bat and the gsc data to monitor high energy transient sources on a daily basis .
all the _ maxi _ gsc light curve data in the current broadband transient monitor is downloaded from the _ maxi _ s public web page .
since all of the light curves are created using an aperture photometry method @xcite , the count rates in the light curves could be contaminated with bright x - ray sources located near the object . due to a relatively poor position resolution of the gsc ,
if other bright x - ray source is located within 2 degrees from the object , there might be an issue in the light curve .
for those sources which could be affected by nearby contaminations , we have a future plan to generate the gsc light curves of the source using the point spread function ( psf ) fit method which should provide a contamination free light curve @xcite using the gsc data . at the current stage
, the sources which are located within 2 degrees from catalog sources in the rosat bright source catalog @xcite with the count rate greater than 1 count s@xmath3 are flagged as possible contaminated sources at the top page of the monitor ( see table [ tbl : src_list ] ) .
the processing time of the current pipeline is limited by the analysis of the bat survey data .
our pipeline is processing the bat survey data at the finest time resolution with the eight energy bands .
although we can process the data in the coarser time resolution and the smaller energy bands to speed up the process , the best spectral and temporal information can be extracted in the current setup for the bat data .
for example , the outputs of our pipeline can provide 8 channel spectral data of the sources every five minutes when high energy sources are in the very bright state .
we have a plan to process the data with multiple available computers or graphical processing unit ( gpu ) to speed up the process .
the current broadband transient monitor contains 106 sources .
however , it is not a difficult task to add sources to the monitor .
we are planning to add sources based on not only the information from various observatories ( e.g. , the astronomer s telegram , the gamma - ray coordinates network ) but also requests from the community .
we would like to thank the anonymous referee for comments and suggestions that materially improved the paper .
this work was supported by jsps grant - in - aid for scientific research ( c ) grant number 25400234 .
abdo , a. a. et al .
2011 , apj , 736 , 131 asai , k. et al .
2015 , pasj , in press barthelmy , s. d. , et al .
2005 , space science reviews , 120 , 3 baumgartner , w. h. , tueller , j. , markwardt , c. b. , skinner , g. k. , barthelmy , s. , mushotzky , r. f. , evans , p. a. , gehrels , n. 2013 , apjs , 207 , 19 esin , a. a. , mcclintock , j. e. , narayan , r. 1997 , apj , 489 , 865 gehrels , n. et al .
2004 , apj , 611 , 2 hiroi , k. et al .
2011 , pasj , 63 , 677 - 689 krimm , h. a. et al .
2013 , apjs , 209 , 14 keck , m.l .
2015 , apj , 806 , 149 matsuoka , m. et al .
2009 , pasj , 61 , 999 mihara , t. et al .
2011 , pasj , 63 , 623 morii , m. et al .
2015 in prep .
nakahira , s. et al .
2010 , pasj , 62 , 27 nakahira , s. et al .
2014 , pasj , 64 , 84 sartore , n. , jourdain , e. , roques , j. p. 2015 , apj , 806 , 193 shidatsu , m. et al .
2011 , pasj , 63 , 785 shidatsu , m. et al .
2011 , pasj , 63 , 803 sugizaki , m. et al .
2011 , pasj , 63s , s644 tueller , j. et al .
2010 , apjs , 186 , 378 voges , w. et al . 1999 ,
a&a , 349 , 389 yoshikawa , a. 2015 , pasj , 67 , 11 name & r.a . & dec .
& possible contaminated source + v709 cas & 7.2036 & 59.2894 & + mrk 348 & 12.1964 & 31.9570 & + gamma cas & 14.1772 & 60.7167 & + smc x-1 & 19.2714 & -73.4433 & t + 2s
0114 + 650 & 19.5112 & 65.2916 & + 4u 0115 + 634 & 19.6330 & 63.7400 & + cc eri & 38.5940 & -43.7960 & + gk per & 52.7993 & 43.9047 & + bq cam & 53.7495 & 53.1732 & + x per & 58.8462 & 31.0458 & + lsv+44 17 & 70.2470 & 44.5300 & + lmc x-4 & 83.2075 & -66.3705 & t + crab nebula pulsar & 83.6332 & 22.0145 & + 1a 0535 + 262 & 84.7274 & 26.3158 & + lmc x-3 & 84.7342 & -64.0823 & t + ngc 2110 & 88.0474 & -7.4562 & + 4u 0614 + 091 & 94.2804 & 9.1369 & + mxb 0656 - 072 & 104.5703 & -7.2105 & + exo 0748 - 676 & 117.1388 & -67.7500 & + vela pulsar & 128.8361 & -45.1764 & + gs 0834 - 430 & 128.9790 & -43.1850 & + vela x-1 & 135.5286 & -40.5547 & + mcg -05 - 23 - 016 & 146.9173 & -30.9489 & + gro j1008 - 57 & 152.4417 & -58.2917 & + ngc 3227 & 155.8774 & 19.8651 & t + mrk 421 & 166.1138 & 38.2088 & + ngc 3516 & 166.6979 & 72.5686 & + xte j1118 + 480 & 169.5450 & 48.0370 & + cen x-3 & 170.3158 & -60.6230 & t + ngc 3783 & 174.7572 & -37.7386 & + 1e 1145.1 - 6141 & 176.8692 & -61.9539 & + ngc
4151 & 182.6357 & 39.4057 & + gx 301 - 2 & 186.6567 & -62.7706 & + 3c 273 & 187.2779 & 2.0524 & + gx 304 - 1 & 195.3217 & -61.6019 & + cen a & 201.3651 & -43.0191 & + 4u 1323 - 619 & 201.6504 & -62.1361 & + ic 4329a & 207.3303 & -30.3094 & + ngc 5506 & 213.3119 & -3.2075 & + psr b1509 - 58 & 228.4813 & -59.1358 & + cir x-1 & 230.1703 & -57.1667 & + h 1538 - 522 & 235.5971 & -52.3861 & + 1e 1547.0 - 5408 & 237.7255 & -54.3066 & + h 1553 - 542 & 239.4512 & -54.4150 & + 4u 1608 - 522 & 243.1792 & -52.4231 & t + sco x-1 & 244.9795 & -15.6402 & + swift j1626.6 - 5156 & 246.6510 & -51.9428 & t + 4u 1624 - 490 & 247.0118 & -49.1985 & + 4u 1626 - 67 & 248.070 & -67.4619 & + 4u 1636 - 536 & 250.2313 & -53.7514 & + gx 340 + 0 & 251.4488 & -45.6111 & + gro j1655 - 40 & 253.5006 & -39.8458 & t + her x-1 & 254.4576 & 35.3424 & t + oao 1657 - 415 & 255.1996 & -41.6731 & t + xte j1701 - 407 & 255.4349 & -40.8583 & t + gx 339 - 4 & 255.7063 & -48.7897 & + gx 349 + 2 & 256.4354 & -36.4231 & t + 4u 1705 - 440 & 257.2270 & -44.1020 & t + 4u 1722 - 30 & 261.8883 & -30.8019 & t + gx 9 + 9 & 262.9342 & -16.9617 & + 4u 1728 - 34 & 262.9892 & -33.8347 & + gx 1 + 4 & 263.0090 & -24.7456 & t + mxb 1730 - 335 & 263.3504 & -33.3877 & + slx 1735 - 269 & 264.5667 & -27.0044 & t + 4u 1735 - 44 & 264.7429 & -44.4500 & + sgr a gal center complex & 266.4168 & -29.0078 & + 2e 1742.9 - 2929 & 266.5229 & -29.5153 & + xte j17464 - 3213 & 266.5650 & -32.2335 & t + gx 3 + 1 & 266.9833 & -26.5636 & + swift j1753.5 - 0127 & 268.3679 & -1.4525 & + gx 5 - 1 & 270.2842 & -25.0792 & t + gx 9 + 1 & 270.3846 & -20.5289 & + sax j1808.4 - 3658 & 272.1150 & -36.9790 & + xte j1810 - 189 & 272.6079 & -19.0700 & + sax j1810.8 - 2609 & 272.6850 & -26.1500 & + gx 13 + 1 & 273.6315 & -17.1574 & + gx 17 + 2 & 274.0058 & -14.0364 & t + 4u 1820 - 30 & 275.9186 & -30.3611 & + 4u 1822 - 371 & 276.4450 & -37.1053 & + ginga 1826 - 24 & 277.3675 & -23.7969 & + xb 1832 - 330 & 278.9338 & -32.9914 & + ser x-1 & 279.9896 & 5.0358 & + ginga 1843 + 00 & 281.4125 & 0.8917 & + 4u 1850 - 086 & 283.2703 & -8.7057 & + v1223 sgr & 283.7593 & -31.1635 & + xte j1855 - 026 & 283.8804 & -2.6067 & + xte j1856 + 053 & 284.1786 & 5.3094 & + hete j1900.1 - 2455 & 285.0360 & -24.9205 & + swift j1910.2 - 0546 & 287.5950 & -5.7990 & + aql x-1 & 287.8167 & 0.5850 & + ss 433 & 287.9565 & 4.9827 & + grs 1915 + 105 & 288.7983 & 10.9456 & + 4u 1916 - 053 & 289.6995 & -5.2381 & + swift j1922.7 - 1716 & 290.6542 & -17.2842 & + 4u 1954 + 31 & 298.9264 & 32.0970 & +
cyg x-1 & 299.5903 & 35.2016 & + cygnus a & 299.8682 & 40.7339 & + v404 cyg & 306.0159 & 33.8672 & + exo 2030 + 375 & 308.0633 & 37.6375 & + cyg x-3 & 308.1074 & 40.9578 & + sax j2103.5 + 4545 & 315.8990 & 45.751 & + ginga 2138 + 56 & 324.8780 & 56.9861 & + cyg x-2 & 326.1717 & 38.3217 & + ngc 7172 & 330.5080 & -31.8698 & t + 4u 2206 + 54 & 331.9843 & 54.5185 & + ii peg & 358.7669 & 28.6337 & + | we present the newly developed broadband transient monitor using the _ swift _ burst alert telescope ( bat ) and the _ maxi _ gas slit camera ( gsc ) data .
our broadband transient monitor monitors high energy transient sources from 2 kev to 200 kev in seven energy bands by combining the bat ( 15 - 200 kev ) and the gsc ( 2 - 20 kev ) data . currently , the daily and the 90-minute ( one orbit ) averaged light curves are available for 106 high energy transient sources .
our broadband transient monitor is available to the public through our web server , http://yoshidalab.mydns.jp/bat_gsc_trans_mon/ , for a wider use by the community .
we discuss the daily sensitivity of our monitor and possible future improvements to our pipeline . |
there are different examples of noncommutative theories , we here concentrate on the case where noncommutativity is described by a constant parameter @xmath5 .
the commutation relations among the coordinates read @xmath6\equiv x^\mu\*x^\nu - x^\nu\ * x^\mu = i\th^{\mu\nu}$ ] , where the star product between functions @xmath7 is given by @xmath8 .
we do not claim that spacetime has exactly this noncommutativity , rather we are interested in investigating a mathematically sound gauge theory based on this easiest noncommutative structure .
general aspects of this noncommutative theory will then probably be in common with more refined choices of @xmath1 .
in particular the choice @xmath9 constant breakes the lorentz group in a spontaneous way ; in a bigger theory we would like to consider @xmath10 ( or the related @xmath11 field ) dynamical and not frozen to a constant value , thus recovering lorentz covariance .
one can also consider gauge theories with @xmath1 nondynamical but frozen to a particular nonconstant value , linear in the coordinates , such that one has a ( kappa ) deformed poincar symmetry , see @xcite .
using seiberg - witten map @xcite , that relates commutative gauge fields to noncommutative ones in such a way that commutative gauge transformations are mapped in nc gauge transformations , one can construct nc gauge theories with arbirary gauge groups @xcite .
these theories are invariant under both commutative and noncommutative gauge transformations . along these lines
noncommutative generalizations of the standard model and gut theories have been studied @xcite .
the sw map and the @xmath12 product allow us to expand these noncommutative actions order by order in @xmath1 and to express them in terms of ordinary commutative fields so that one can then study the physics properties of these @xmath1-expanded commutative actions , see for ex .
it turns out that given a commutative ym theory , sw map and commutative / noncommutative gauge invariance are in general not enough in order to single out a unique noncommutative generalization of the original ym theory .
one can follow different criteria in order to select a specific noncommutative generalization .
we here focus on a classical analysis , in particular imposing the constraint that the noncommutative generalization of the standard model should be compatible with noncommutative gut theories .
another issue would be to single out a noncommutative sm or gut that is well behaved at the quantum level .
we refer to the problems relative to renormalization , see for ex . @xcite . on the other hand chiral gauge anomalies
are absent in these models @xcite . in this
talk , following @xcite , we present a general study of the ambiguities that appear when constructing ncym theories .
we then see that at first order in @xmath1 there is no ambiguity in @xmath0 ncym theory .
in particular no triple gauge bosons coupling of the kind @xmath13 is present .
we further study the noncommutative sm compatible with @xmath0 .
we next study the reality , hermiticity , charge conjugation , parity and time reversal properties of the sw map and of @xmath1-expanded ncym theories .
this constraints the possible freedom in the choice of a `` good '' sw map . in @xcite the @xmath3 properties of ncqed
were studied assuming the usual @xmath14 and @xmath15 transformations also for noncommutative fields .
we here show that the usual @xmath3 transformation on commutative spinors and nonabelian gauge potentials imply , via sw map , the same @xmath3 transformations for the noncommutative spinors and gauge potentials .
we also see that @xmath16 is always a symmetry of noncommutative actions . in @xcite @xmath16
is studied more axiomatically .
the reality property of the sw map is then used to analyze the difference between the sm in @xcite and the gut inspired sm proposed here .
it is a basic one , and can be studied also in a qed model . while in @xcite , and in general in the literature , left and right handed components of a noncommutative spinor field are built with the same sw map , we here use and advocate a different choice : if noncommutative left handed fermions are built with the @xmath17 sw map then their right handed companions should be built with the @xmath18 sw map ; this implies that both noncommutative @xmath19 and @xmath20 are built with the @xmath17 sw map .
in other words , with this choice , noncommutativity does not distinguish between a left handed fermion and a left handed antifermion , but does distinguish between fermions with different chirality .
this appears to be the only choice compatible with gut theories .
consider an ordinary `` commutative '' ym action with gauge group @xmath21 , and one fermion multiplet , @xmath22 this action is gauge invariant under @xmath23 where @xmath24 is the representation of @xmath21 determined by @xmath25 . following @xcite the noncommutative generalization of this action
is given by = d^4x -12 g^2^ t^r(_^ ) + i [ action1multiplet ] where the noncommutative field strength @xmath26 is defined by @xmath27\ , .
$ ] the covariant derivative is given by d_= _ - i _ ( a _ ) .
[ covder ] the action ( [ action1multiplet ] ) is invariant under the noncommutative gauge transformations = i _ ( ) , a_= _
the fields @xmath28 , @xmath29 and @xmath30 are functions of the commutative fields @xmath31 and the noncommutativity parameter @xmath1 via the sw map @xcite . at first order in @xmath1
we have @xmath32 & = & a_\xi + \frac{1}{4 } \theta^{\mu\nu}\{a_\nu,{\partial}_\mu a_\xi\ } + \frac{1}{4 } \theta^{\mu\nu}\{f_{\mu\xi},a_\nu\ } + \mathcal{o}(\theta^2 ) \label{swa}\\ \widehat\lambda[\lambda , a , \th ] & = & \lambda + \frac{1}{4 } \theta^{\mu\nu}\{{\partial}_\mu \lambda , a_\nu\}+ \mathcal{o}(\theta^2)\label{swlambda}~\\ \widehat \psi[\psi , a,\th ] & = & \psi + \frac{1}{2 } \theta^{\mu\nu}\rho_\psi(a_\nu){\partial}_\mu\psi + \frac{i}{8}\theta^{\mu\nu}[\rho_\psi(a_\mu ) , \rho_\psi(a_\nu ) ] \psi + \mathcal{o}(\theta^2 ) \label{swpsi}\end{aligned}\ ] ] in terms of the commutative fields the action ( [ action1multiplet ] ) is also invariant under the ordinary gauge transformation @xmath33 $ ] , @xmath34 . in ( [ action1multiplet ] )
the information on the gauge group @xmath21 is through the dependence of the noncommutative fields on the commutative ones .
the commutative gauge potential @xmath35 and gauge parameter @xmath36 are valued in the @xmath21 lie algebra , @xmath37 ; and from ( [ swa ] ) , ( [ swlambda ] ) it follows that @xmath28 and @xmath38 are valued in the universal enveloping algebra of the @xmath21 lie algebra .
however , due to the sw map , the degrees of freedom of @xmath28 are the same as that of @xmath35 . similarly to @xmath28 , also @xmath26 is valued in the universal enveloping algebra of @xmath21
. now expression ( [ action1multiplet ] ) is ambiguous because in @xmath39 we have not specified the representation @xmath40 .
we can render explicit the ambiguity in ( [ action1multiplet ] ) by writing t^r(_^)= _ c_((_)(^ ) ) [ ambiguity ] where the sum is extended over all unitary irreducible and inequivalent representations @xmath41 of @xmath21 . the real coefficients @xmath42 parametrize the ambiguity in ( [ ambiguity ] ) .
they are constrained by requiring that in the commutative limit , @xmath43 , ( [ ambiguity ] ) becomes the correctly normalized commutative gauge kinetic term .
the ambiguity ( [ ambiguity ] ) in the action ( [ action1multiplet ] ) can also be studied by expanding ( [ ambiguity ] ) in terms of the commutative fields @xmath44 . at first order in @xmath1
we have _ gauge&= & - d^4x _
f^a_f^a _ + & & + ( _ c_d_^abc ) ^4d^4x^ 14f^a _ f^b _ f^c -f^a _ f^b _ f^c [ traccia ] where d_^abc((t^a)\{(t^b),(t^c ) } ) = a()(t^a\{t^b , t^c})d^abc[dterm ] . here
@xmath45 denotes the fundamental representation , and we are using that the completely symmetric @xmath46 tensor in the representation @xmath41 is proportional to the @xmath47 one defined by the fundamental representation .
in particular for all simple lie groups , except @xmath48 with @xmath49 , we have @xmath50 for any representation @xmath41 .
thus from ( [ traccia ] ) we see that at first order in @xmath1 the ambiguity ( [ ambiguity ] ) is present just for @xmath48 lie groups . among the possible representations that one can choose in ( [ ambiguity ] )
there are two natural ones .
the fermion representation and the adjoint representation .
the adjoint representation is particularly appealing if we just have a pure gauge action , then , since only the structure constants appear in the commutative gauge kinetic term @xmath51 , a possible choice is indeed to consider only the adjoint representation .
this is a minimal choice in the sense that in this case only structure constants enter ( [ ambiguity ] ) .
( it can be shown @xcite that in this case the gauge action is even in @xmath1 ) .
if we also have matter fields then from ( [ covder ] ) we see that we must consider the particle representation @xmath52 given by the multiplet @xmath25 ( and inherited by @xmath29 ) . in ( [ ambiguity ] ) one could then make the minimal choice of selecting just the @xmath52 representation . along the lines of the above ncym theories framework we now examine the @xmath0 , the @xmath2 and the standard model noncommutative gauge theories .
* noncommutative * we consider only one fermion generation : the @xmath53-dimensional spinor representation of @xmath0 usually denoted @xmath54 ( no relevant new effects appear considering all three families ) .
we write the left handed multiplet as _ l^+=(u^i , d^i ,- u^c_i , d^c_i,,e^-,e^+ , -^c)_l [ ml ] where @xmath55 is the @xmath56 color index and @xmath57 is the charge conjugate of the neutrino particle @xmath58 ( not present in the standard model ) .
the gauge and fermion sector of noncommutative @xmath0 is then simply obtained by replacing @xmath29 with @xmath59 in ( [ action1multiplet ] ) .
notice that no linear term in @xmath1 , i.e. no cubic term in @xmath60 can appear .
this is so because @xmath0 is anomaly free : @xmath50 forall @xmath41 . in other words , at first order in @xmath1 , noncommutative @xmath0 gauge theory is unique .
* noncommutative * the fermionic sector of @xmath2 has the @xmath61 multiplet that transforms in the @xmath62 of @xmath2 and the @xmath63 multiplet that transforms according to the @xmath64 of @xmath2 . in this case
we expect that the adjoint , the @xmath65 and the @xmath64 representations enter in ( [ ambiguity ] ) . in principle
one can consider the coefficients @xmath66 , i.e. while the @xmath67 fermion rep .
is @xmath68 , in ( [ ambiguity ] ) the weights @xmath42 of the @xmath65 and the @xmath64 can possibly be not the same .
it turns out that only if @xmath66 then @xmath69 in ( [ traccia ] ) .
we see that , already at first order in the noncommutativity parameter @xmath1 , noncommutative @xmath2 gauge theory is not uniquely determined by the gauge coupling constant @xmath70 , but also by the value of @xmath71 . thus @xmath2 is _ not _ a truly unified theory in a noncommutative setting .
it is tempting to set @xmath72 so that exactly the fermion representation @xmath73 enters ( [ traccia ] ) .
we then have @xmath74 , ( however this relation is not protected by symmetries ) .
* ( gut inspired ) noncommutative standard model @xmath75 * one proceeds similarly for the sm gauge group . the full ambiguity of the gauge kinetic term is given in @xcite .
about the fermion kinetic term , the fermion vector @xmath76 is constructed from @xmath77 .
the covariant derivative is as in ( [ covder ] ) , with @xmath78 and with @xmath79 , where @xmath80 are the generators of @xmath81 .
the fermion kinetic term is then as in ( [ action1multiplet ] ) .
this standard model is built using only left handed fermions and antifermions .
we call it gut inspired because its noncommutative structure can be embedded in @xmath0 gut .
indeed @xmath82 and @xmath83 differ just by the extra neutrino @xmath57 ; moreover under an infinitesimal gauge transformation @xmath30 , all fermions in @xmath82 transform with @xmath30 on the left .
this gut inspired standard model differs from the one considered in @xcite ; indeed here we started from the chiral vector @xmath82 , while there the vector @xmath84 is considered . in the commutative case @xmath85 but in the noncommutative case
( see later ) this is no more true : @xmath86 ; if we change @xmath1 into @xmath18 in the right handed sector of @xmath87 , then the two expressions coincide .
finally it is a natural choice to consider in the sm gauge kinetic term only the adjoint rep . and the fermion rep .
, we then have that at first order in @xmath4 there are no modifications to the sm gauge kinetic term .
this is so because the fermion rep .
is anomaly free : @xmath88 and because for @xmath89 the adjoint rep .
is trivial . *
higgs sectors@xmath75 * while the noncommutative higgs kinetic and potential terms are given by ( d_)^d^ + ^2 ^ - ^ ^ , a noncommutative version of the sm and gut yukawa terms is not straighforward and requires the introduction of the hybrid seiberg - witten maps @xmath90 on fermions .
a typical noncommutative yukawa term then reads ^ ^^___h + herm .
[ hsw ] where @xmath91 . under an infinitesimal @xmath81 gauge transformation @xmath36 ,
@xmath92 transforms as @xmath93 .
we see that in the hybrid sw map @xmath30 appears both on the left and on the right of the fermions , moreover the representation of @xmath30 is inherited from the higgs and fermions that sandwich @xmath92 .
the yukawa term ( [ hsw ] ) is thus invariant under noncommutative gauge transformations .
of course in the @xmath94 limit we recover the usual gauge transformation for the leptons .
an explicit formula for the hybrid sw map at first order in @xmath1 is in @xcite .
the yukawa terms ( [ hsw ] ) differ from those studied in @xcite .
there the hybrid sw map is considered on @xmath95 , in particular there @xmath96 is not invariant under @xmath56 gauge transformations ( and this implies that in @xcite gluons couple directly to the higgs field ) . the higgs sector in the @xmath0 and @xmath2 models can be constructed with similar techniques @xcite .
from ( [ swa ] ) we see that if @xmath35 is hermitian then @xmath97 is also hermitian
. actually , to all orders in @xmath1 , @xmath28 and @xmath30 can be chosen hermitian if @xmath35 and @xmath36 are hermitian .
otherwise stated , sw map can be chosen to be compatible with hermitian conjugation .
compatibility of sw map with complex conjugation reads , = ^^ [ conj ] where @xmath98 $ ] denotes the sw map of @xmath25 constructed with the representation @xmath52 of the potential @xmath35 , and the sw map of the complex conjugate spinor @xmath99 is defined by [ 30 ] where @xmath100 is the representation conjugate to @xmath52 we have @xmath101 and , since @xmath102 , @xmath103 are real , we have @xmath104 . ] .
notice that in ( [ 30 ] ) the noncommutativity parameter @xmath4 appears with opposite sign w.r.t .
the @xmath4 in the sw map of @xmath25 .
similarly to ( [ conj ] ) we have @xmath105 and @xmath106 where @xmath107 $ ] and @xmath108 $ ] .
the proof of ( [ conj ] ) relies on showing that the sw differential equations @xcite ( obtained by requiring that gauge equivalence classes of the gauge theory with noncommutativity @xmath109 , correspond to gauge equivalence classes of the gauge theory with noncommutativity @xmath1 ) are themselves compatible with complex conjugation @xcite .
one proceeds similarly for the case of hermitian conjugation . *
noncommutativity and chirality@xmath75 * we can now discuss a further ambiguity of noncommutative gauge theories , and resolve it by requiring compatibility with grand unified theories .
for simplicity we consider noncommutative qed .
let @xmath110 be a @xmath111-component dirac spinor , and decompose it into its weil spinors @xmath19 and @xmath112 .
their charge conjugate spinors are @xmath113 and @xmath114 .
consider the noncommutative left - handed spinor @xmath115~,$ ] we then have the @xmath116 choice = [ choice ] for the right handed one . in the literature the choice @xmath17 is usually considered so that for the @xmath111-component dirac spinor @xmath110 we can write @xmath117 $ ] , @xmath118 .
we here advocate the opposite choice ( @xmath18 ) in ( [ choice ] ) . indeed from ( [ conj ] )
we have that @xmath119 and therefore = = [ equiv ] so that with the @xmath18 choice in ( [ choice ] ) , both left handed fermions @xmath120 , @xmath121 are associated with @xmath1 while the right handed ones @xmath122 , @xmath123 are associated with @xmath18 . in gut theories we have multiplets of definite chirality and
therefore this is the natural choice to consider in this setting .
these observations allow us to compare qed@xmath124 with @xmath125 , the two different qed theories obtained with the two different @xmath126 choices ( [ choice ] ) .
this difference immediately extends to the fermion kinetic terms of nonabelian gauge theories and allows us to compare the ncsm discussed in @xcite with the present gut compatible one .
we have ( up to gauge kinetic terms ) @xmath127 where the gut inspired @xmath125 is obtained using the left handed spinor @xmath128 so that @xmath129 $ ] . now from @xmath119 and from @xmath130 matrix algebra we have @xmath131 , where we have emphasized that we are using the @xmath18 convention in the sw map by writing @xmath132 instead of @xmath133 .
we conclude that in order to obtain qed@xmath134 from qed@xmath124 we just need to change @xmath1 into @xmath18 in the right handed fermion sector of qed@xmath124 .
we have studied the @xmath14 and @xmath15 symmetry properties of ncym actions where @xmath1 transforms under @xmath14 and @xmath15 as a field strenght .
viceversa , _ if we keep @xmath1 fixed _ under @xmath14 and @xmath15 transformations , we in general have that ncym theories _ break _ @xmath14 and
@xmath15 symmetries .
finally a @xmath89 field strenght is invariant under the combined @xmath16 transformation , and therefore @xmath1 does not change .
this implies that @xmath16 is always a symmetry of ncym actions .
j. madore , s. schraml , p. schupp and j. wess , eur .
j. c * 16 * , 161 ( 2000 ) [ hep - th/0001203 ] .
+ b. jurco , s. schraml , p. schupp and j. wess , eur .
j. c * 17 * , 521 ( 2000 ) [ hep - th/0006246 ] .
b. jurco , l. moller , s. schraml , p. schupp and j. wess , eur .
j. c * 21 * , 383 ( 2001 ) [ hep - th/0104153 ] .
x. calmet , b. jurco , p. schupp , j. wess and m. wohlgenannt , eur .
j. c * 23 * ( 2002 ) 363 [ arxiv : hep - ph/0111115 ] .
p. aschieri , b. jurco , p. schupp and j. wess , nucl .
b * 651 * ( 2003 ) 45 [ arxiv : hep - th/0205214 ] .
p. schupp , j. trampetic , j. wess and g. raffelt , `` the photon neutrino interaction in non - commutative gauge field theory and astrophysical bounds , '' arxiv : hep - ph/0212292 . m. buric and v. radovanovic , jhep * 0402 * ( 2004 ) 040 [ arxiv : hep - th/0401103 ] . f. brandt , c. p. martin and f. r. ruiz , jhep * 0307 * ( 2003 ) 068 [ arxiv : hep - th/0307292 ] .
m. chaichian , k. nishijima and a. tureanu , phys .
b * 568 * , 146 ( 2003 ) [ arxiv : hep - th/0209008 ] + l. alvarez - gaume and m. a. vazquez - mozo , nucl .
b * 668 * ( 2003 ) 293 [ arxiv : hep - th/0305093 ] + n. mahajan , phys .
b * 569 * ( 2003 ) 85 [ arxiv : hep - th/0305105 ] . | noncommutative generalizations of yang - mills theories using seiberg - witten map are in general not unique .
we study these ambiguities and see that @xmath0 gut , at first order in the noncommutativity parameter @xmath1 , is unique and therefore is a truly unified theory , while @xmath2 is not .
we then present the noncommutative standard model compatible with @xmath0 gut .
we next study the reality , hermiticity and @xmath3 properties of the seiberg - witten map and of these noncommutative actions at all orders in @xmath4 .
this allows to compare the standard model discussed in @xcite with the present gut inspired one . |
spinning objects have historically been interesting subjects to study .
the spin reversal of the rattleback @xcite ( also called a celt or wobblestone ) and the behavior of the tippe top are typical examples . in the latter case ,
when a truncated sphere with a cylindrical stem , a so - called ` tippe top ' , is spun sufficiently rapidly on a table with its stem up , it will flip over and rotate on its stem .
this inversion phenomenon has fascinated physicists and has been studied for over a century @xcite . in the present paper
we revisit and study this very classical problem from a different perspective .
recently the riddle of spinning eggs has been resolved by moffatt and shimomura [ ms ] @xcite .
they discovered that if an axisymmetric body , such as a hard - boiled egg , is spun sufficiently rapidly , a ` _ gyroscopic balance _ ' condition ( gbc ) holds and that under this condition the governing equations of the system are much simplified .
in particular , they derived a first - order ordinary differential equation ( ode ) for @xmath8 , the angle between the axis of symmetry and the vertical axis , and showed for the case of a prolate spheroid that the axis of symmetry indeed rises from the horizontal to the vertical .
then the spinning behavior of egg - shaped axisymmetric bodies , whose cross sections are described by several models of oval curves , was studied under the gbc by one of the present authors @xcite .
the tippe top is also an axisymmetric body and shows the similar behavior as the spinning egg .
then one may ask : does the gbc also hold for the tippe top ?
if so , how is it related to the inversion phenomenon of the tippe top ?
in the first half of this paper we analyze the spinning motion of the tippe top in terms of the gbc .
actually the gbc is not satisfied initially for the tippe top , contrary to the case of the spinning egg .
the difference comes from how we start to spin the object : we spin the tippe top with its stem up , in other words , with its symmetry axis vertical while the egg is spun with its symmetry axis horizontal . in this paper
we perform our analysis taking an eccentric sphere version of the tippe top instead of a commercially available one , a truncated sphere with a cylindrical stem . in order to examine the gbc of the tippe top more closely , we introduce a variable @xmath2 so that @xmath3 corresponds to the gbc , and study the behavior of @xmath2 .
numerical analysis shows that for the tippe tops which will turn over , the variable @xmath2 , starting from a large positive value @xmath9 , soon takes negative values and fluctuates around a negative but small value @xmath10 such that @xmath11 .
thus for these tippe tops , the gbc , which is not satisfied initially , will soon be realized but approximately . on the other hand , in the case of the tippe tops which will not turn over , @xmath2 remains positive around @xmath9 or changes from positive @xmath9 to negative values and then back to positive values close to @xmath9 again .
we find that the behavior of @xmath2 is closely related to the inversion phenomenon of the tippe top .
once @xmath2 fluctuates around the value @xmath12 , the system becomes unstable and starts to turn over . under the gbc
the governing equations for the tippe top are much simplified and we obtain a first - order ode for @xmath8 , which has the same form as the one derived by ms for the spinning egg . then , this equation for @xmath8 and the geometry lead to the classification of tippe tops into _ three _ groups , depending on the values of @xmath13 and @xmath14 , where @xmath15 and @xmath16 are two principal moments of inertia , and @xmath17 is the distance from the center of sphere to the center of mass and @xmath18 is the radius of sphere .
the tippe tops of group i never flip over however large a spin they are given .
those of group ii show a complete inversion and the tippe tops of group iii tend to turn over up to a certain inclination angle @xmath4 such that @xmath5 , when they are spun sufficiently rapidly .
this classification of tippe tops into three groups and its classificatory criteria totally coincide with those obtained by hugenholtz @xcite and leutwyler @xcite , both of whom resorted to completely different arguments and methods . in the latter half of this paper we study the steady states for spinning motion of the tippe top and
examine their stability ( or instability ) .
it is well understood that the main source for the tippe top inversion is sliding friction @xcite , which depends on the slip velocity @xmath0 of the contact point between the tippe top and a table . often used
is coulomb friction ( see eq.([coulomb ] ) ) .
in fact , coulomb friction is practical when @xmath19 is away from zero , but it is undefined for @xmath20 .
however , we learn that at the steady state of the tippe top , the slip velocity @xmath0 necessarily vanishes . in order to facilitate a linear stability analysis of steady states and also to study the motion of the tippe top as realistically as possible ,
we adopt in our analysis a modified version of coulomb friction ( see eq.([friction ] ) ) , which is continuous in @xmath0 and vanishes at @xmath1 .
actually the steady states of the tippe top and their stability were analyzed by ebenfeld and scheck [ es ] @xcite , who assumed a similar frictional force which is continuous at @xmath1 .
they used the total energy of the spinning top as a liapunov function .
the steady states were found as solutions of constant energy . and the stability or instability of these states were judged by examining whether the liapunov function assumes a minimum or a maximum at these states .
also recently , bou - rabee , marsden and romero [ bmr ] @xcite analyzed the tippe top inversion as a dissipation - induced instability and , using the modified maxwell - bloch equations and an energy - momentum argument , they gave criteria for the stability of the non - inverted and inverted states of the tippe top .
we take a different approach to this problem .
first , in order to find the steady states for spinning motion of the tippe top , we follow the method used by moffatt , shimomura and branicki [ msb ] for the case of spinning spheroids @xcite
. then the stability of these steady states is examined as follows : once a steady state is known , the system is perturbed around the steady state . particularly we focus our attention on the variable @xmath8 , which is perturbed to @xmath21 , where @xmath22 is a value at the steady state and @xmath23 is a small quantity .
using the equations of motion , we obtain , under the linear approximation , a first - order ode for @xmath23 of the form , @xmath24 , where @xmath25 is expressed by the values of dynamical variables at the steady state . thus the change of @xmath23 is governed by the sign of @xmath25 .
if @xmath25 is positive ( negative ) , @xmath26 will increase ( decrease ) with time .
therefore , we conclude that when @xmath25 is negative ( positive ) , then the state is stable ( unstable ) . using this new and rather intuitive criterion we argue about the stability of the steady states in terms of the initial spin velocity @xmath6 given at the position near @xmath27 .
we observe that our results on the stability of the steady states are consistent with ones obtained by es and msb .
then we obtain a critical value @xmath7 of the initial spin which is required for the tippe top of group ii to flip over up to the completely inverted position at @xmath28 .
finally we confirm by simulation our results on the relation between the initial spin @xmath6 and the stability of the steady states .
the paper is organized as follows : in sec . 2 we explain the notation and geometry used in this paper , and give all the necessary equations for the analysis of the spinning motion of the tippe top .
in sec . 3 we discuss about the gbc and its relevance to the inversion phenomenon of the tippe top .
we also show that the assumption of the gbc leads to the classification of tippe tops into three groups . then in sec .
4 we study the steady states for the spinning motion of the tippe top and examine their stability .
is devoted to a summary and discussion .
in addition , we present four appendices . in appendix
a , the equations of motion which are used to analyze the spinning motion of the tippe top are enumerated . in appendix b , it is shown that intermediate steady states for the tippe tops of group ii and group iii are stable when an initial spin @xmath29 falls in a certain range . in appendix c
we demonstrate that our stability criterion for the steady state is equivalent to the one obtained by es . and
finally , in appendix d , we show that our results on the stability of the vertical spin states are consistent with the criteria derived by bmr .
a commercially available tippe top is usually a truncated sphere with a cylindrical stem .
instead we perform our analysis taking a loaded ( eccentric ) sphere version of the tippe top .
the center of mass is off center by a distance @xmath17 .
there are no qualitative differences between the two .
but if applied to the case of a commercial tippe top with a stem , our assertions would be valid up to the point when the stem touched the table surface . a loaded sphere ( eccentric ) version of the tippe top .
the center of mass @xmath30 is off center ( @xmath31 ) by distance @xmath17 .
the tippe top spins on a horizontal table with point of contact @xmath32 .
its axis of symmetry , @xmath33 , and the vertical axis , @xmath34 , define a plane @xmath35 , which precesses about @xmath34 with angular velocity @xmath36 .
@xmath37 is a rotating frame of reference with @xmath38 horizontal in the plane @xmath35 .
the height of @xmath30 above the table is @xmath39 , where @xmath18 is the radius .
the position vector of @xmath32 from @xmath30 is @xmath40 , where @xmath41 and @xmath42 .
, scaledwidth=60.0% ] fig .
[ loadedsphere ] shows the geometry .
an axisymmetric tippe top spins on a horizontal table with point of contact @xmath32 .
we will work in a rotating frame of reference @xmath37 , where the center of mass is at the origin , @xmath30 .
the center @xmath31 of the sphere with radius @xmath18 is at a distance @xmath17 from the origin .
the symmetry axis of the tippe top , @xmath33 , and the vertical axis , @xmath34 , define a plane @xmath35 , which precesses about @xmath34 with angular velocity @xmath36 .
let @xmath43 be the euler angles of the body relative to @xmath44 .
then we have @xmath45 , where the dot represents differentiation with respect to time , and @xmath8 is the angle between @xmath34 and @xmath33 .
we choose the horizontal axis @xmath38 in the plane @xmath35 and thus @xmath46 is vertical to @xmath35 and inward . in a rotating frame of reference @xmath47 , where @xmath48 is in the plane @xmath35 and perpendicular to the symmetry axis @xmath33 and where @xmath49 coincides with @xmath46 , the tippe top spins about @xmath33 with the rate @xmath50 .
since @xmath51 is expressed as @xmath52 in the frame @xmath47 , the angular velocity of the tippe top , @xmath53 , is given by @xmath54 .
here @xmath55 , @xmath56 , and @xmath57 are unit vectors along @xmath48 , @xmath49 , and @xmath33 , respectively , @xmath58 is given by @xmath59 .
the @xmath48 and @xmath49 are not body - fixed axes but are principal axes , so that the angular momentum , @xmath60 , is expressed by @xmath61 , where @xmath62 are the principal moments of inertia at @xmath30 . using the perpendicular axis theorem and the parallel axis theorem
, we see that @xmath63 for any axisymmetric density distribution .
the coordinate system @xmath47 is obtained from the frame @xmath37 by rotating the latter about the @xmath46 ( @xmath49 ) axis through the angle @xmath8 .
hence , in the rotating frame @xmath37 , @xmath53 and @xmath60 have components @xmath64 respectively .
the evolution of @xmath60 is governed by euler s equation @xmath65 where @xmath66 is the position vector of the contact point @xmath32 from @xmath30 , @xmath67 is the normal reaction at @xmath32 , @xmath68 , with @xmath69 being of order @xmath70 , the weight , and @xmath71 is the frictional force at @xmath32 .
we consider only _ the situation in which the tippe top is always in contact with the table _ throughout the motion .
since the point @xmath32 lies in the plane @xmath35 , @xmath66 has components @xmath72 , which are given by @xmath73 where @xmath74 is the height of @xmath30 above the table .
the components of ( [ euler ] ) are expressed , respectively , as @xmath75 in terms of @xmath8 , @xmath76 , and @xmath77 the above equations are rewritten as @xmath78 now it is easily seen from ( [ angulareqs ] ) , ( [ eulerx ] ) and ( [ eulerz ] ) that there exists an exact constant of motion , @xmath79 which is valid irrespective of the reaction force @xmath80 at the contact point @xmath32 , in other words , whether or not slipping occurs .
this so - called jellett s constant " @xcite is typical for the tippe top whose portion of the surface in contact with the table is spherical .
the velocity , @xmath81 , of the contact point @xmath32 with respect to the center of mass @xmath30 is given by @xmath82 , and thus has components , @xmath83 the center of mass @xmath30 is not stationary .
let @xmath84 represent the velocity of @xmath30 , then the slip velocity of the contact point @xmath32 , @xmath85 , is @xmath86 since @xmath87 , we have @xmath88 as was expected .
the equation of motion for the center of mass @xmath30 is given by @xmath89 where @xmath90 is the mass of the tippe top and @xmath91 is the force of gravity . in components ,
eq.([eqcm ] ) reads @xmath92 since @xmath93 , eq.([eqcmz ] ) gives @xmath94 which shows that the normal force @xmath69 is of order @xmath95 when @xmath96 .
we need an information on the frictional force @xmath97 .
it is well understood that the sliding friction is the main source for the tippe top inversion @xcite .
so we will ignore other possible frictions , such as , rolling friction @xcite and rotational friction which is due to pure rotation about a vertical axis .
concerning the sliding friction , often used is a coulomb law , which states that @xmath98 where @xmath99 is a coefficient of friction .
another possibility is a viscous friction law , which states that the friction is linearly related to @xmath0 .
coulomb friction is practical when @xmath19 is away from zero but it is undefined at @xmath100 .
the slip velocity of the contact point @xmath32 necessarily vanishes at the steady state of the tippe top . in order to study the motion of the tippe top as realistically as possible and also to facilitate a linear stability analysis of steady states
, we modify the expression of coulomb friction ( [ coulomb ] ) as @xmath101 so that @xmath97 is continuous in @xmath0 and vanishes at @xmath102 . here
we choose @xmath103 as a sufficiently small number with dimensions of velocity .
note that @xmath88 and thus the @xmath104-component of @xmath97 is 0 .
this completes the presentation of all the necessary equations for the analysis of the motion of tippe tops .
we enumerate all these equations in appendix a. we need further the initial conditions .
when we play with a tippe top , we usually give it a rapid spin with its axis of symmetry nearly vertical .
so let us choose the following initial conditions for @xmath8 and other angular velocities : @xmath105 we take @xmath106 rad and @xmath107 .
recall that the spin @xmath58 is given by @xmath59 , and thus we have @xmath108 . as for the initial condition for the velocity of the center of mass @xmath30
, we take @xmath109 since we usually do not give a large translational motion to the tippe top at the beginning . with the above initial conditions ( [ initialrotation ] ) and ( [ initialtrans ] )
, we analyze the behaviors of the tippe top using three angular ( [ euleromega]-[eulern ] ) and three translational ( [ eqcmx]-[eqcmz ] ) equations of motion , together with the knowledge of the frictional force , a modified version of the coulomb law ( [ friction ] ) , and the velocities ( [ vpx]-[vpz ] ) and ( [ velp ] ) . when we perform simulations we use the adaptive runge - kutta method .
we define a variable @xmath2 as @xmath110 in terms of @xmath2 , the @xmath111- and @xmath104- components of @xmath60 in ( [ angulareqs ] ) and jellett s constant @xmath112 , ( [ jellettconstant ] ) , are expressed , respectively , as @xmath113 the condition @xmath3 has been introduced by ms @xcite in their analysis of spinning hard - boiled eggs , and referred to as the gbc .
they discovered that the gbc , @xmath3 , is approximately satisfied for the spinning egg and , using this gbc , they resolved a long standing riddle : when a hard - boiled egg is spun sufficiently rapidly on a table with its axis of symmetry horizontal , the axis will rise from the horizontal to the vertical .
we outline how ms found the gbc for the spinning egg @xcite . the system of the spinning egg obeys essentially the same equations of motion as the case of the tippe top , to be specific , eqs .
( [ euler ] ) and ( [ eqcm ] ) .
the @xmath114-component of ( [ euler ] ) for the spinning egg is given by ( [ eulertheta ] ) , with the factor , @xmath115 , being replaced by @xmath116 . because the secular change of @xmath8 is slow and
thus @xmath117 , the term @xmath118 can be neglected .
furthermore , in a situation where @xmath119 is sufficiently large so that the terms involving @xmath76 in ( [ eulertheta ] ) dominate the terms @xmath120 and @xmath121 , eq .
( [ eulertheta ] ) is reduced , in leading order , to @xmath122 .
hence , for @xmath123 , we arrive at the condition @xmath124 .
the tippe top shows the similar behavior as the spinning egg .
then one may ask : does the gbc also hold for the tippe top ? we will show that the answer is partly no " and partly yes " . partly no " means that the gbc is not satisfied initially .
tippe tops are usually spun with @xmath125 , @xmath126 , and large @xmath127 and , therefore , @xmath128 is large , from which we find that @xmath129 is large[multiblock footnote omitted ] .
thus the gbc does not hold at the beginning .
however , we will see later that the gbc does approximately hold whenever the tippe top rises , which is the meaning of partly yes " . in fact , the argument of ms to derive the gbc for the spinning egg can also be applied to the tippe top .
thus in a situation where @xmath76 is sufficiently large and for @xmath123 , the gbc is expected to be satisfied . on the other hand , in the case of the spinning egg
, the gbc is approximately satisfied initially .
we start to spin an egg with its symmetry axis horizontal , that is , with @xmath130 , @xmath131 and large @xmath132 .
hence we find @xmath133 and @xmath134 for the spinning egg .
we emphasize that the variable @xmath2 initially takes a large positive value for the tippe top .
but our numerical analysis will show that when a tippe top turns over , @xmath2 soon makes a rapid transition from large positive values to negative values and starts to oscillate about a small negative value . before proceeding with a discussion of this transition of @xmath2 ,
let us consider the consequences when the gbc is exactly satisfied for the tippe top .
in a situation where @xmath76 is sufficiently large and @xmath8 is not in the vicinity of 0 or @xmath135 , the gbc is realized for the tippe top .
let us consider the case that the exact gbc , @xmath136 , is satisfied for the tippe top .
then , we have @xmath137 from ( [ jellettconstant2 ] ) , and @xmath138 from the second equation in ( [ angulareqs2 ] ) .
if the angular velocity @xmath76 around the vertical axis is reduced and , therefore , @xmath139 decreases , eq.([gbcjellett ] ) tells us that the height @xmath74 of the center of mass from the table increases since @xmath112 is a constant , which means the turning over of the tippe top . differentiating both sides of ( [ gbcjellett ] ) by time and using ( [ height.b ] ) and ( [ eulerz ] )
, we obtain a first - order ode for @xmath8 , @xmath140 we assume also that the @xmath114-component of @xmath141 , the translational velocity of the center of mass @xmath30 , in ( [ velp ] ) is negligible in the first approximation as compared with that of @xmath81 , and we set @xmath142 . we see that numerical simulation supports this assumption
then , one can use eq .
( [ vpy ] ) and the gbc to eliminate @xmath77 and @xmath76 , and obtain @xmath143 as only a function of the dynamical variable @xmath8 as follows : @xmath144 since the frictional force @xmath145 is proportional to @xmath143 , we obtain from eqs.([equfortheta]-[vpsecond ] ) , @xmath146 with a _ positive _ proportional coefficient and @xmath147 equation ( [ equtheta ] ) implies that the change of @xmath8 is governed by the sign of @xmath148 .
if @xmath148 is positive ( negative ) , then @xmath8 will increase ( decrease ) with time .
therefore a close examination of the behavior of @xmath148 as a function of @xmath8 will be important . ) to a renormalization group equation which appears in quantum field theories for critical phenomena and high energy physics is emphasized in sec .
we observe from ( [ vpytilde ] ) that @xmath149 at @xmath150 and @xmath135 , since @xmath151 at these angles .
moreover , @xmath148 may vanish at an other angle , which is given by solving @xmath152 equation ( [ vpzero ] ) has a solution for @xmath8 if @xmath153 or @xmath154 and no solution otherwise . accordingly , tippe tops are classified into three groups , depending on the values of @xmath13 and @xmath14 : group i with @xmath153 ; group ii with @xmath155 ; and group iii with @xmath154 .
we now examine the behaviors of tippe tops belonging to each group . [ cols="^,^ " , ] from these numerical analyses we see that the behavior of @xmath2 is closely related to the inversion phenomenon of the tippe top .
as the tippe top turns over , simulation shows that @xmath2 becomes small ( in the sense @xmath156 ) and takes values close to @xmath157 , which implies that the relation ( [ gbcjellett ] ) is approximately satisfied .
conversely , when the relation ( [ gbcjellett ] ) holds , it means that the center of mass of the tippe top goes up as @xmath139 decreases .
in sec.3.2 we have studied the behaviors of the spinning tippe top when the gyroscopic balance condition @xmath3 is exactly satisfied .
the situation corresponds to the one in which the tippe top is given an infinitely large initial spin .
actually the initial spin given to the tippe top is finite and we know empirically that a tippe top with a small spin is stable and does not turn over .
we will now consider how large an initial spin should be for the tippe top to turn over .
for that purpose we will study the steady states of the tippe top and examine their stability . actually the steady states ( or the asymptotic states ) of the tippe top and their stability
were analyzed by ebenfeld and scheck [ es ] @xcite .
they used the total energy of the spinning top as a liapunov function .
then the steady states were found as solutions of constant energy .
the stability or instability of these states was determined by examining whether the liapunov function assumes a minimum or a maximum at these states under the constraint of jellett s constant .
the tippe top inversion was also analyzed recently by bou - rabee , marsden and romero [ bmr ] @xcite as a dissipation - induced instability .
bmr used the modified maxwell - bloch equations and an energy - momentum argument to determine the stability of the non - inverted and inverted states of the tippe top . here
we take a different approach to this problem . and
we discuss the stability of the steady states in terms of the initial spin velocity @xmath77 given at the non - inverted position @xmath27 .
recently , moffatt , shimomura and branicki [ msb ] made a linear stability analysis of the spinning motion of spheroids @xcite .
they identified the steady states , and then discussed their stability and found the critical angular velocity needed for the rise of the body . in order to find the steady states of the spinning tippe top , we adopt the method taken by msb for the case of spheroids .
but for the stability analysis of the steady states , we develop a new stability criterion which is different from the ones used by es , bmr and msb .
our approach to the stability problem of the tippe top is as follows .
once a steady state is known , the system is perturbed around the steady state .
particularly we focus our attention on the variable @xmath8 , which is perturbed to @xmath158 where @xmath22 is a value at the steady state and @xmath23 is a small quantity . using the equations of motion , we obtain , under the linear approximation , a first - order ode for @xmath23 of the following form : @xmath159 where @xmath160 and @xmath161 are values taken at the steady state . equation ( [ eqforstability ] ) implies that the change of @xmath23 is governed by the sign of the function @xmath162 . if @xmath162 is positive ( negative ) , @xmath26 will increase ( decrease ) with time .
therefore , we conclude that _ when @xmath162 is negative ( positive ) , then the state is stable ( unstable)_. this is the criterion for stability of the steady state , which we will use in this paper .
the stability criterion in this work is derived from an intuitive analysis of the equations of motion .
we check in appendices c and d that they are consistent with those derived by es and bmr which are based on mathematically rigorous methods .
superficially the above criterion ( [ eqforstability ] ) seems quite different from the one used by es @xcite , but actually we have found that both are equivalent and , therefore , our results are consistent with theirs . in appendix c
we will show the equivalence of both criteria and that the stability conditions of the steady states which we will obtain coincide with the ones found by es .
after all , es utilized the total energy ( an integral form ) @xcite , while we will use equations of motion ( differential forms ) .
the criterion ( [ eqforstability ] ) for the stability of the tippe top also seems different from the ones used by bmr @xcite , which were derived from the tippe top modified maxwell - bloch equations . in order to obtain the stability criteria ,
both bmr and we linearize equations of motion about the steady states and use sliding friction , which is assumed to be an analytic function of the slip velocity , as the main mechanism behind tippe top inversion .
thus it is well expected that both criteria lead to the consistent results on the stability of the non - inverted and inverted states .
( the stability of the intermediate states have not been analyzed yet by means of the modified maxwell - bloch equations ) . in appendix
d we will show that the expressions of the criteria provided in bmr become more transparent when they are rewritten in terms of the parameters and classification criteria used in this paper , and that they lead to the same stability conditions for the vertical spinning states which will be obtained later by using the criterion ( [ eqforstability ] ) .
besides , although bmr did not mentioned , the classification of tippe tops into three groups , group i , ii , and iii , is shown to be possible through the close examination of the criteria in bmr . the steady states of the spinning motion of the tippe top are obtained from the equations of motion ( [ euleromega]-[eulern ] ) and ( [ eqcmx]-[eqcmz ] ) by setting @xmath163 @xcite . since we assume that the sliding friction ( [ friction ] ) , i.e. , a modified version of coulomb law ) is non - analytic at @xmath102 and a nonlinear friction law that would not appear in the linear approximation @xcite .
] , is the only frictional force present , the energy equation @xmath164 shows that @xmath102 and @xmath165 at the steady states @xcite .
thus we obtain for the steady states of the tippe top , @xmath166 where @xmath167 and the velocity equations ( [ vpx]-[vpy ] ) and ( [ velp ] ) have been used . the solutions for eqs.([uoxzero]-[vpyzero ] )
are : \i ) vertical spin state at @xmath168 : @xmath169 which is a spinning state about the axis of symmetry with the center of mass below the sphere center .
\ii ) vertical spin state at @xmath28 : @xmath170 which is an overturned spinning state about the axis of symmetry with the center of mass above the sphere center .
\iii ) intermediate states : @xmath171 the elimination of @xmath77 from ( [ intera ] ) and ( [ interb ] ) gives @xmath172 the necessary ( but not sufficient ) condition for the existence of such states is @xmath173 recall ( [ vpzero ] ) which was used for the classification of tippe tops into three groups in sec . 3.2 . thus
, intermediate states may exist at @xmath174 for the tippe top of group i ( @xmath153 ) , at @xmath8 between 0 and @xmath135 for group ii ( @xmath155 ) , and at @xmath175 for group iii ( @xmath154 ) .
there appear , in total , three categories of steady states for a loaded sphere version of the tippe top .
the turnover of the tippe top is associated with the effect of the sliding friction ( with a coefficient @xmath99 ) at the point of contact @xmath32 . near the steady states ,
we know that @xmath176 , which is equivalent to the situation where @xmath177 .
thus for the stability analysis of the steady states , we consider the limiting case of @xmath178 @xcite .
since we expect @xmath179 near the steady states , we have @xmath180 and @xmath181 . eq.([vpy ] ) shows @xmath182 , which leads to @xmath183 and , hence , @xmath184 .
then ( [ eqcmy ] ) gives @xmath185 , and @xmath186 from ( [ vpx ] ) and ( [ velp ] ) , and thus we have @xmath187 and @xmath188 .
the above order estimation in @xmath99 near the steady states leads to the primary balance in ( [ eulertheta ] ) which holds at leading order in @xmath99 @xcite , @xmath189 note that with a sufficiently large @xmath76 , eq.([stabilityomega ] ) reduces to @xmath190 , the gbc .
the angle @xmath8 is perturbed from @xmath150 , and we take @xmath191 . in the linear approximation
we may take @xmath192 , since eq.([eulern ] ) implies that @xmath193 is quadratic in small quantities ( note @xmath194 ) . with @xmath195 , the primary balance ( [ stabilityomega ] ) gives @xmath196 in this approximation @xmath76 is also a constant .
then eq.([euleromega ] ) gives @xmath197 and we may take @xmath198 since @xmath199 .
hence , we require for the stability at @xmath168 @xmath200 using the expressions of both `` @xmath201 '' and `` @xmath202 '' solutions for @xmath76 in ( [ omegaat0 ] ) , the above condition is rewritten as @xmath203 which gives @xmath204 it is easily seen that the requirement ( [ zerostability2 ] ) is always satisfied for any spin velocity @xmath77 by the tippe top of group i ( @xmath205 . as for the tippe top of group ii or iii with @xmath206 , the requirement ( [ zerostability2 ] ) is rewritten as @xmath207 the stability of the vertical spin state at @xmath168 is summarized as follows : for the tippe top of group i with @xmath208 , the spinning state at @xmath168 is stable for any spin @xmath77 , while for the tippe top of group ii or iii with @xmath206 we require @xmath209 for its stability . in other words ,
the tippe top of group ii or iii becomes unstable at @xmath168 if @xmath210 a similar analysis can be made for the stability of the spinning state at @xmath28 .
now put @xmath211 with @xmath212 .
again we may take @xmath192 , but note that @xmath77 may be negative near @xmath213 . with @xmath214 , the primary balance ( [ stabilityomega ] )
gives @xmath215 in order for @xmath76 to have a real solution , we require @xmath216 for @xmath217 , the spin is insufficient to overcome the effect of gravity and the orientation becomes unstable @xcite . with @xmath218 ,
( [ euleromega ] ) gives @xmath219 and we may take , @xmath220 hence we require for the stability at @xmath28 , @xmath221 using the expressions of both `` @xmath201 '' and `` @xmath202 '' solutions for @xmath76 in ( [ omegaatpi ] ) , the above condition gives @xmath222 first , the requirement ( [ pistability2 ] ) is never satisfied by the tippe top of group iii @xmath223 .
so the tippe top of group iii is unstable at @xmath28 .
actually it never turns over to the position with @xmath28 . for the tippe top of group
i or ii which satisfies @xmath224 , the requirement ( [ pistability2 ] ) becomes @xmath225 note that @xmath226 .
the stability of the vertical spin state at @xmath28 is summarized as follows : for the tippe top of group iii ( @xmath227 ) , the spinning state at @xmath213 is unstable for any spin @xmath77 , while for the tippe top of group i or ii with @xmath228 , the state at @xmath28 is stable if @xmath229 we have learned in sec.4.2.1 that the spinning state of group i at @xmath150 is stable .
we also know from the discussion in sec.4.1 that the intermediate steady states of group i , if they exist , must occur at @xmath230 .
this implies that the spinning motion of group i near @xmath150 does not shift to a possible intermediate steady state . on the other hand ,
the tippe tops of group ii and iii become unstable at @xmath150 when they are spun with a sufficiently large initial spin @xmath231 , where @xmath232 is given by ( [ zerostability4 ] ) , and they will start to turn over . here
we are interested in the intermediate steady states of the tippe top which are reached from the initial spinning position near @xmath150 .
therefore , in this subsection , we focus on the possible steady states only for the tippe tops of group ii and iii , and examine their stability . the jellett s constant given by ( [ jellettconstant ] ) or ( [ jellettconstant2 ] ) is rewritten as @xmath233 now eqs.([intera ] ) and ( [ interb ] ) and the above expression of j completely determine the intermediate steady states .
they are derived by solving @xmath234= \bigl\{(\cos\theta-\frac{a}{r})^2+\frac{a}{c}\sin^2\theta\bigr\}^2 ~,\label{eqforintermediate}\ ] ] where @xmath235 define the following function : @xmath236 where @xmath237 and @xmath238 then , eq.([eqforintermediate ] ) is rewritten as @xmath239 since @xmath240 , we obtain @xmath241 ^ 2}(\frac{a}{c}-1)~ , \label{derivg}\\ f''(x ) & = & \frac{2}{[f_1(x)]^3}\bigl\ { \bigl([f_1(x)]^2+(\frac{a}{c}-1 ) \sqrt{f_2(x ) } \bigr ) ^2 + 3 [ f_1(x)]^4 \bigr\ } > 0~. \label{derivderivg}\end{aligned}\ ] ] the condition for the initial spin @xmath231 means @xmath242 . using ( [ zerostability4 ] ) , we find @xmath243 , which leads to @xmath244 .
so we are looking for solutions of @xmath245 with @xmath244 .
( i)group ii @xmath246 when @xmath247 , @xmath248 and @xmath249 is a monotonically decreasing function for @xmath250 .
hence , there is one and only one solution of @xmath245 at @xmath251 between @xmath252 and 1 , provided @xmath253 .
otherwise , there is no solution , which means that there exists no intermediate steady state . expressing @xmath112 with the initial spin at @xmath150 as @xmath254 , we find that the condition @xmath255 gives @xmath256 thus , in the case @xmath257 , one intermediate steady state exists at @xmath251 , provided that @xmath258 we know @xmath259 from ( [ derivderivg ] ) , and so @xmath249 is concave upward for @xmath250 .
if @xmath260 , then @xmath249 may have a local minimum at a certain @xmath261 between @xmath252 and 1 .
recall that we are looking for the steady states which are reached from the position near @xmath168 and that the requirement for this is @xmath244 .
hence , for the existence of such a steady state we need @xmath262 the first condition @xmath263 gives @xmath264 and the second one @xmath265 leads to @xmath266 .
some tippe tops of group ii with @xmath267 satisfy @xmath268 as well as the conditions ( [ condsmallac ] ) , and thus @xmath269 . for such tippe tops , the corresponding @xmath249 has a local minimum between @xmath270 and 1 , where @xmath270 is a solution of @xmath271 .
these tippe tops , therefore , have one intermediate steady state at @xmath251 between @xmath272 and @xmath270 when the condition @xmath273 is satisfied .
see the discussion of case ( c ) in fig.[thetafgroupii ] . for the tippe tops of group ii with @xmath274
, there exists no intermediate state .
we will see later , in the discussion of case ( d ) in fig.[thetafgroupii ] , that these tippe tops will turn over to @xmath28 once given a spin @xmath275 , since @xmath276 and , hence , @xmath277 for these tops .
( ii)group iii @xmath278 since @xmath279 should be positive , the allowed region of @xmath261 is @xmath280 with @xmath281 .
eq.([derivg ] ) together with @xmath282 shows that @xmath249 is a monotonically decreasing function for @xmath280 .
note that @xmath249 positively diverges when @xmath261 approaches @xmath283 from larger @xmath261 .
hence , once @xmath244 is satisfied , @xmath245 has one and only one solution at @xmath251 such that @xmath284 . in other words
, one intermediate steady state always exists at @xmath285 between 0 and @xmath286 for the tippe top of group iii , if the condition @xmath275 is satisfied .
when @xmath287 gets larger , the angle @xmath22 gets closer to @xmath4 but never crosses @xmath4 . in order for @xmath22 to reach @xmath4 , @xmath287 should be infinite .
now we know that there exists an intermediate steady state for the tippe top of group ii with property @xmath288 , when @xmath289 satisfies @xmath290 . also there is an intermediate steady state for the tippe top of group iii with @xmath291 if @xmath292 .
let ( @xmath293 ) represent such a steady state so that ( @xmath293 ) are related by eqs.([intera ] ) and ( [ interb ] ) , and suppose this state to be perturbed to @xmath294 noting that @xmath295 and @xmath296 , we find that the perturbed state satisfies @xmath297 where @xmath298 ^ 2+\bigl(\frac{a}{c } -1 \bigr ) \sqrt{f_2(x_s ) } ~. \label{dxs}\ ] ] the details of the derivation of ( [ eqthetastability2 ] ) are given in appendix b. if @xmath299 , then @xmath300 . also when @xmath269 , we find that @xmath301 is still positive ( see appendix b ) .
thus we observe from ( [ eqthetastability2 ] ) that @xmath302 with a negative constant at the intermediate steady state , which means that this state is indeed stable .
finally it is emphasized that the spinning state of the tippe top of group i is stable at @xmath150 and the top will not turn over from the position near @xmath150 . on the other hand , the tippe top of group iii , when given a sufficiently large spin near the position @xmath150 , will tend to turn over and approach the steady state at @xmath22 but never up to the inverted position at @xmath28 .
the tippe top of group ii will turn over to the inverted position at @xmath213 when it is given a sufficient initial spin .
let us estimate the critical value @xmath7 so that the spinning top with @xmath303 reaches the inverted position . recall that jellett s constant ( [ jellettconstant3 ] ) is invariant during the turnover from @xmath150 to @xmath213 . from the relation @xmath304
, we obtain @xmath305 we already know that we need @xmath306 for the stability at @xmath213 , where @xmath307 is given in ( [ pistability4 ] ) .
thus we find @xmath308 also from the instability condition of the tippe top of group ii at @xmath150 , we need @xmath231 , where @xmath232 is given by ( [ zerostability4 ] ) . hence the condition for the tippe top of group ii to turn over up to @xmath28 is that the initial spin @xmath289 should be larger than both @xmath309 and @xmath232 .
in fact , we observe @xmath310 for the tippe top with @xmath311 , while @xmath312 for the tippe top with @xmath313 , where @xmath314 is given by ( [ rc ] ) .
therefore , we obtain @xmath315 we now study the time evolution of the inclination angle @xmath8 from a spinning position near @xmath150 .
simulations are made with various values of @xmath13 and @xmath14 , changing the input parameters @xmath15 and @xmath17 .
other input parameters are the same as those given in ( [ inputparameter ] ) .
initial conditions are @xmath316 rad , @xmath317 , and @xmath318 , and the initial value of the spin velocity @xmath6 is varied .
since we have chosen a very small @xmath319 , we may consider @xmath6 as @xmath289 .
the asymptotic value @xmath320 as a function of the initial spin velocity @xmath6 for tippe tops of group ii with various values of @xmath13 and @xmath14 ; ( a ) the one with @xmath321 and @xmath322 ; the others have @xmath323 but different @xmath14 such as ( b ) @xmath322 , ( c ) @xmath324 and ( d ) @xmath325 . ]
figure [ thetafgroupii ] shows the asymptotic ( final ) angle of inclination , @xmath320 , as a function of @xmath6 for several types of tippe tops of group ii with different values of @xmath13 and @xmath14 ; ( a ) the one with @xmath321 and @xmath322 ; the others have @xmath323 but different @xmath14 such as ( b ) @xmath322 , ( c ) @xmath324 , and ( d ) @xmath325 .
the asymptotic angle @xmath320 may be 0 or @xmath135 , or @xmath22 , the angle of a possible intermediate steady state .
the symbols @xmath326 , @xmath327 , @xmath328 and @xmath329 represent the results for the tippe tops ( a ) , ( b ) , ( c ) and ( d ) , respectively , and the thin solid curves ( a ) , ( b ) and ( c ) are the trajectories obtained by solving ( [ eqforintermediate ] ) .
we observe that the numerical results fall on the predicted curves .
the values of @xmath330 , in units of rad / sec , for the tops ( a ) , ( b ) , ( c ) and ( d ) are 34.4(62.9 ) , 42.1(54.5 ) , 45.7(49.4 ) and 51.4(44.4 ) , respectively . in each case
we see that the spinning state near @xmath150 is stable when @xmath331 .
once @xmath6 gets larger than @xmath232 , the state becomes unstable and the tippe top turns over up to the asymptotic angle @xmath320 . for
the tippe tops ( a ) and ( b ) the values of @xmath320 grow with @xmath6 from 0 to @xmath135 . on the other hand ,
the tippe top ( c ) satisfies @xmath269 with @xmath332 , and thus the intermediate steady state exists only at @xmath333 with @xmath334 , where @xmath335 is a solution of @xmath336 .
we find @xmath337 .
thus when @xmath6 gets larger than @xmath232 for the case of the tippe top ( c ) , the asymptotic angle @xmath320 jumps from 0 to @xmath335 . when @xmath338 , @xmath339 for the tops ( a ) , ( b ) and ( c ) .
in the case of the tippe top ( d ) , we find @xmath340 and thus @xmath341 , which leads to @xmath342 .
therefore , there is no intermediate steady state , and the asymptotic angle @xmath320 is 0 or @xmath135 depending on @xmath343 .
the asymptotic value @xmath320 as a function of the initial spin velocity @xmath6 for tippe tops of group iii : ( a ) with @xmath344 and @xmath345 ; and ( b ) with @xmath344 and @xmath322 . ] we plot in fig.[thetafgroupiii ] the asymptotic angle @xmath320 as a function of @xmath6 for the tippe tops of group iii ; ( a ) with @xmath346 and @xmath347 and ( b ) with @xmath346 and @xmath348 .
the symbols @xmath326 and @xmath327 represent the results of simulation for the tippe tops ( a ) and ( b ) , respectively , and the thin solid curves ( a ) and ( b ) are the trajectories obtained by solving ( [ eqforintermediate ] ) .
we observe again that the numerical results on @xmath320 for both tops ( a ) and ( b ) fall on the predicted curves .
the values of @xmath232 for the tops ( a ) and ( b ) are 13.3 and 23.5 rad / sec , respectively . in both cases
the spinning position near @xmath168 is stable when @xmath6 is below @xmath232 . above @xmath232 ,
the value of @xmath320 grows with @xmath6 and approaches the fixed point @xmath4 .
the values of @xmath4 for the tops ( a ) and ( b ) are 1.67 and 2.21 rad , respectively . for simulations we have used a modified version of the coulomb friction @xmath97 given in ( [ friction ] ) .
the value @xmath320 is not affected by the strength of the coefficient @xmath99 .
the strength of @xmath99 instead has an effect on the rate of rising of the tippe top .
if we use another form than ( [ friction ] ) for the sliding friction , and moreover , it is expressed as a continuous function of @xmath0 and vanishes at @xmath102 , then we still expect that we get the same numerical results on @xmath320 _ vs. _ @xmath6 as shown in fig.[thetafgroupii ] and fig.[thetafgroupiii ] .
this is due to the observation that the numerical value @xmath320 has fallen on the predicted curves which are derived from ( [ eqforintermediate ] ) and that we have obtained ( [ eqforintermediate ] ) using the property of @xmath97 which vanishes at the steady states together with @xmath0 .
the time evolution of the angle @xmath8 for a tippe top of group ii from a spinning position near @xmath168 . ]
figure [ thetaevolution ] shows the time evolution of the inclination angle @xmath8 for a tippe top of group ii from a spinning position near @xmath168 for various values of the initial spin velocity @xmath6 .
input parameters and initial conditions are the same as before and we take @xmath321 and @xmath322 .
the asymptotic angles @xmath320 which will be reached are 0 , 0.92 , 1.67 , 2.49 , @xmath135 and @xmath135 rad for @xmath34930 , 40 , 50 , 60 , 70 and 80 rad / sec , respectively .
simulations with a modified version of the coulomb friction ( [ friction ] ) show that the larger value of @xmath6 is given , the faster the rate of rising becomes .
we have examined an inversion phenomenon of the spinning tippe top , focusing our attention on its relevance to the gyroscopic balance condition ( gbc ) , which was discovered by moffatt and shimomura in the study of the spinning motion of a hard - boiled egg . in order to analyze the gbc in detail for the case of the tippe top
, we introduce a variable @xmath2 given by ( [ xi ] ) so that @xmath3 corresponds to the gbc , and study the behavior of @xmath2 .
contrary to the case of the spinning egg , the gbc is not satisfied initially for the tippe top .
the simulation shows that , starting from a large positive value @xmath9 , the variable @xmath2 for the tippe tops which rise , soon fluctuates around a negative but small value @xmath10 such that @xmath11 .
thus for these tippe tops , the gbc , though it is not fulfilled initially , will soon be satisfied approximately .
once @xmath2 fluctuates around the value @xmath12 , these tops become unstable and start to turn over . on the other hand , in the case of the tippe tops which do not turn over
, @xmath2 remains positive around @xmath9 or changes from positive @xmath9 to negative values and then back to positive values close to @xmath9 again . under the gbc
the governing equations for the tippe top are much simplified and , together with the geometry of the tippe top , we obtain a first - order ode for @xmath8 in the following form @xcite ( see ( [ equfortheta ] ) or ( [ equtheta ] ) ) : @xmath350 it is noted that this equation has a remarkable resemblance to the renormalization group ( rg ) equation for the effective coupling constant @xmath351 , @xmath352 which appears in quantum field theories for critical phenomena @xcite and high energy physics @xcite . here in ( [ rgequation ] ) , @xmath353 is expressed as @xmath354 with a dimensionless scale parameter @xmath355 .
provided that @xmath356 has a zero at @xmath357 , we find that , if @xmath358 , then @xmath359 as @xmath360 ( @xmath361 ) , and while if @xmath362 , @xmath359 as @xmath363 ( @xmath364 )
. the limiting value @xmath365 of @xmath366 is known as the ultraviolet ( infrared ) fixed point in the former ( latter ) case .
similarity between the two equations , ( [ equthetalast ] ) and ( [ rgequation ] ) , and the notion of the rg equation brought us to a consequence that tippe tops are classified into three groups , depending on the values of @xmath13 and @xmath14
. a resemblance of eq.([eqforstability ] ) to the rg equation also gave us a hint that eq.([eqforstability ] ) might serve as a criterion for stability of the steady state in sec .
the criterion ( 4.2 ) is a first - order ode for the ( perturbed ) inclination angle @xmath23 , and the results derived from this criterion coincide with those by es and bmr which are obtained by mathematically rigorous methods .
the key ingredients in the process of arriving at this first - order ode are the order estimation in @xmath99 near the steady states and an intuitive analysis of the equations of motion .
the criterion ( 4.2 ) can also be applied to the stability analysis of other spinning objects .
in fact we have applied ( [ eqforstability ] ) to the spinning motion of spheroids ( prolate and oblate ) which was recently examined in detail by msb @xcite , and we have obtained consistent results with theirs .
finally we have assumed , in the present work , a modified version of coulomb law ( [ friction ] ) for the sliding friction , since coulomb friction ( [ coulomb ] ) is non - analytic and undefined at @xmath100 . on the other hand ,
cohen used coulomb friction in his pioneering work on the tippe top @xcite , and analyzed its spinning motion numerically for the first time .
he reported the result of a sample simulation in fig.5 of his paper @xcite .
the coulomb friction is realistic provided that @xmath19 is away from zero , but its application to the spinning motion of the tippe top is very delicate . near steady states ( i.e. , near @xmath150 or @xmath135 or @xmath22 )
, @xmath0 almost vanishes ( see , for example , fig.[travelocity ] ( a ) and ( b ) ) .
and there the @xmath111- and @xmath114-components of @xmath367 are changing signs rapidly and moreover non - analytically , and so are the components of friction , @xmath368 and @xmath145 .
coulomb friction may not be adequate to be applied to such a situation .
in fact , kane and levinson @xcite argued against the work of cohen , because it did not include adequate provisions for transitions from sliding to rolling and vice versa .
they reanalyzed the simulation of cohen , assuming coulomb law for sliding friction , but also providing an algorithm that rolling begins when @xmath369 ( with @xmath370 1m / sec ) is satisfied , together with another algorithm for the transition from rolling to sliding .
they found that a transition from sliding to rolling occurs soon after the motion has begun and that values of @xmath8 remain below 0.077 rad thereafter . or @xcite adopted a hybrid friction law adding viscous friction , which is linearly related to @xmath0 , to coulomb friction .
other frictional forces such as the one which is due to pure rotation about the normal at the point of contact might have some effect . after all it is safe to say that we have understood general features of the tippe top inversion .
but it would be not until we have had thorough knowledge of frictional force that we completely understood the inversion phenomena of the tippe top . _ and
yet , it flips over_. we thank tsuneo uematsu for valuable information on the spinning egg and the tippe top .
we also thank yutaka shimomura for introducing us to the paper @xcite and for helpful discussions . *
we enumerate the equations of motion which are used to analyze the spinning motion of the tippe top : @xmath371
there exists an intermediate steady state for the tippe top of group ii with property @xmath288 , when an initial spin @xmath289 satisfies @xmath266 .
there is also an intermediate steady state for the tippe top of group iii if @xmath231 . in this appendix
we show that these steady states are stable . near
the steady states the primary balance condition ( [ stabilityomega ] ) holds at leading order in @xmath99 . differentiating both sides of ( [ stabilityomega ] ) with respect to @xmath353 , we obtain @xmath372 using ( [ euleromega ] ) and ( [ eulern ] ) , and eliminating @xmath373 and @xmath193 , we find @xmath374 let ( @xmath293 ) represent an intermediate steady state so that @xmath160 , @xmath161 and @xmath22 are related by ( [ intera ] ) and ( [ interb ] ) , and suppose this state to be perturbed to @xmath294 since @xmath295 and @xmath296 , the perturbed state satisfies @xmath375 where @xmath376 at leading order in @xmath99 , we have @xmath377 ( recall @xmath378 ) , and thus we obtain from ( [ vpy ] ) , @xmath379 now we expect that the perturbed state still satisfies the primary balance condition ( [ stabilityomega ] ) , since @xmath380 and @xmath368 are @xmath381 .
then a variation around the steady state gives @xmath382 where @xmath383 is given by ( [ ss ] ) . also taking a variation of jellett constant ( [ jellettconstant3 ] ) around the steady state ( and then , of course , we have @xmath384 )
, we obtain @xmath385 from ( [ varia ] ) and ( [ varib ] ) , @xmath386 and @xmath387 are expressed in terms of @xmath23 as @xmath388 inserting these expressions into ( [ deltafy ] ) , and then we obtain from ( [ eqthetastability1 ] ) @xmath389 where @xmath298 ^ 2+\bigl(\frac{a}{c } -1 \bigr ) \sqrt{f_2(x_s ) } ~,\ ] ] and @xmath390 , and eqs.([f1 ] ) and ( [ f2 ] ) have been used . if @xmath391 , then @xmath300 .
also when @xmath269 , @xmath301 is still positive , which is explained as follows : the expression of ( [ derivg ] ) shows that the function @xmath392 is related to @xmath393 as @xmath394 ^ 2}d(x)~.\ ] ] when @xmath269 , an intermediate steady state at @xmath395 exists provided @xmath263 and @xmath265 . at that point @xmath396 is negative , and thus @xmath301 is positive .
ebenfeld and scheck @xcite analyzed the stability of the spinning tippe top using the total energy as a liapunov function and gave the stability criteria for the steady states .
we take a different approach to this stability problem .
first the system is perturbed around the steady state .
then , using the equations of motion and under the linear approximation , we obtain a first - order ode for @xmath23 of the form given in ( [ eqforstability ] ) .
we make use of this equation and give a different stability criterion . in this appendix
we show that both approaches are equivalent and thus they lead to the same conclusions on the stability conditions of the steady states .
es wrote the total energy of the spinning top as the sum of two terms ( es-(33 ) ) ) of ref.@xcite as es-(@xmath397 ) .
the jellett constant @xmath355 defined by es is related to our @xmath112 as @xmath398 . ]
@xmath399 the second of which contains all the terms that will vanish at the steady states , while the first depends on @xmath400 and jellett s constant @xmath112 . in terms of the parameters used in this paper , @xmath401 and @xmath402
are expressed as follows : @xmath403}{rg(\eta_3 ) } \biggr\}^2~ , \label{e2}\end{aligned}\ ] ] with @xmath404 note that es set @xmath405 .
the condition @xmath406 together with @xmath407 leads to the three solutions of the steady states : ( i ) vertical spin state at @xmath150 ( [ steadythetazero ] ) , ( ii ) vertical spin state at @xmath213 ( [ steadythetapi ] ) , and ( iii ) intermediate states ( [ eqforintermediate ] ) , or equivalently , ( [ intera]-[interb ] ) .
it is recalled that we have obtained these solutions starting from equations of motion . at these steady states
@xmath402 vanishes . for intermediate steady states ,
the factor @xmath408/rg(\eta_3 ) \}$ ] in ( [ e2 ] ) reduces to zero , due to ( [ intera]-[interb ] ) and jellett s constant given in ( [ jellettconstant3 ] ) .
now we show that the criterion , eq.([eqforstability ] ) , for the stability of the steady states is equivalent to the one derived by es @xcite . for the stability analysis of the steady states ,
the order estimation in @xmath99 near the steady states is important , which has been pointed out by msb in their work on the linear stability analysis of the spinning motion of spheroids @xcite . as explained at the beginning of sec .
4.2 , near the steady states we have @xmath179 , @xmath409 , @xmath410 , and @xmath188 .
since @xmath402 is already @xmath411 ( recall that it vanishes at the steady states ) , we have @xmath412 , while @xmath413 . thus near the steady states ,
the energy equation ( [ energyeq ] ) is written at leading order in @xmath99 as @xmath414 suppose the steady states to be perturbed to @xmath415 . since @xmath416
, we have @xmath417 , and @xmath418 is expanded as @xmath419 where we have used the fact @xmath420 .
meanwhile @xmath421 is shown to be expressed as @xmath422 actually we have already obtained the expressions ( [ vpythetazero ] ) and ( [ vpythetapi ] ) for @xmath423 near the steady states at @xmath150 and @xmath213 , respectively . also near the intermediate steady states , @xmath386 and @xmath387 are expressed in terms of @xmath23 as ( [ deltans ] ) and ( [ deltaomegas ] ) , respectively , and thus we obtain ( [ vrotpydelta ] ) . now using eqs.([energyeqleading])-([vrotpydelta ] )
we find @xmath424 which means that we can identify @xmath162 in ( [ eqforstability ] ) as @xmath425 hence we conclude that the following assertions are equivalent : _ a steady state is stable ( unstable ) _ @xmath426 _
@xmath162 is negative ( positive ) _
@xmath427 is positive ( negative)_. in fact , es showed that if the quantity ( es-(39 ) ) with the upper sign is positive , then @xmath427 is positive at @xmath428 and the non - inverted rotating motion is liapunov stable . on the other hand , starting from the equations of motion we derived @xmath162 and obtained the condition ( [ zerostability2 ] ) for the stability of the rotating motion at @xmath428 .
it is easily seen that the statement that the quantity ( es-(39 ) ) with the upper sign is positive is equivalent to the inequality given in ( [ zerostability2 ] ) , once we know that jellett s constant at @xmath428 is given by @xmath429 .
similarly , if the quantity ( es-(39 ) ) with the lower sign is positive , then @xmath427 is positive at @xmath430 and the completely inverted rotating motion is liapunov stable .
the condition that the quantity ( es-(39 ) ) with the lower sign is positive is equivalent to the inequality given in ( [ pistability2 ] ) . note , this time , @xmath431 . as for the intermediate steady state @xmath432
, es stated that if the steady state exists and the quantity ( es-(40 ) ) is negative , then @xmath427 is positive and the state is liapunov stable .
in sec.4.2.3 we have shown that the stability of the intermediate steady state is determined by the sign of @xmath301 given in ( [ dxs ] ) .
now it is interesting to note that @xmath301 is related to ( es-(40 ) ) as follows : @xmath433 ^ 2}{ac^2}\times ( { \rm es}.(40))~.\ ] ] hence the condition that the quantity ( es-(40 ) ) is negative is equivalent to @xmath434 .
we have seen in sec.4.2.3 that there exists an intermediate steady state for the tippe top of group ii and also of group iii .
( we have not considered a possible intermediate steady state for group i , since such a state , even if it exists , can not be reached from the initial spinning position near @xmath168 . ) for these steady states , we have shown , in appendix b , that @xmath301 is positive and , therefore , the states are stable .
recently bou - rabee , marsden and romero [ bmr ] treated tippe top inversion as a dissipation - induced instability .
they showed that the modified maxwell - bloch ( mmb ) equations are a normal form for tippe top inversion and , using the mmb equations and an energy - momentum argument , they gave criteria for the stability on the non - inverted and inverted states of the tippe top @xcite .
although we have not explored the connections between the mmb equations and the first - order ode ( [ eqforstability ] ) for @xmath23 , we show in appendix d that our results on the stability of the vertical spin states are consistent with the criteria provided by bmr . actually , rewritten in terms of dimensional parameters and classification criteria used in this paper ,
the expressions of those criteria become more transparent and they lead to the same stability conditions as ours for the vertical spinning states . besides , although bmr did not mentioned , the classification of tippe tops into three groups , group i , ii , and iii , according to the behaviors of spinning motion , is possible from the close examination of those criteria .
bmr used the moments of inertia defined as the ones about the principal axes _
attached to the center of sphere instead of the center of mass_. the correspondence between the parameters used by bmr and ones in this paper are as follows : @xmath435 where @xmath436 is the spin rate of the initially standing equilibrium solution ( we added a subscript bmr to distinguish from our @xmath76 ) , and the dimensionless bmr s jellett " constant , @xmath437 , is restricted to have a certain value , i.e. , @xmath438 .
also bmr expressed the vector from the center of sphere to the center of mass @xmath439 ( in the bmr notation @xmath440 ) as @xmath441 , where @xmath442 is a unit vector along the symmetry axis . using the tippe top modified maxwell - bloch equations
, bmr obtained the stability criteria for the non - inverted state which are given by the three inequalities in ( bmr-(5.3 ) ) ) of ref.@xcite as bmr-(@xmath397 ) . ] .
they took @xmath443 ( upward ) in ( bmr-(5.3 ) ) .
since the non - inverted state has the center of mass below the center of sphere , we have @xmath444 , and thus @xmath445 . the first inequality of ( bmr-(5.3 ) )
is rewritten as @xmath446 , which is always satisfied .
apart from some irrelevant positive constants , the second and third inequalities are expressed , respectively , as @xmath447 ^ 2c}\bigl(1-\frac{a}{r } \bigr)\frac{a}{c}+\bigl(1-\frac{a}{r } \bigr)^5 \frac{\nu^2}{\sigma^2}-\frac{a}{c}+\bigl(1-\frac{a}{r } \bigr)>0~ , \label{2ndineq } \\ & & -\bigl\ { \frac{a}{c}- \bigl(1-\frac{a}{r } \bigr ) -\frac{mga}{[n(\theta=0)]^2c } \bigl(1-\frac{a}{r } \bigr)^2\bigr\}>0~. \label{3rdineq}\end{aligned}\ ] ] from these inequalities , we find : * in the case @xmath448 , i.e. , for the tippe top of group i , the above inequalities are always satisfied .
in other words , the non - inverted states ( @xmath150 ) of group i are always stable .
* in the case @xmath449 , i.e. , for the tippe tops of group ii or iii , the inequality ( [ 3rdineq ] ) is satisfied if @xmath450 ^ 2<\frac{mga}{c\ { \frac{a}{c } -(1-\frac{a}{r})\}}\bigl(1- \frac{a}{r } \bigr)^2 , \label{d4}\ ] ] which is the same requirement given in ( [ zerostability3 ] ) for the stability of the tippe top of group ii or iii .
note that the inequality ( [ 2ndineq ] ) is automatically satisfied when both @xmath449 and inequality ( [ 3rdineq ] ) hold .
the inequalities ( bmr-(5.3 ) ) , which were derived as the stability criteria for the non - inverted state , can also be used for the stability criteria for the inverted state , but with some replacements .
since @xmath443 ( upward ) , the inverted state has the center of mass above the center of sphere .
thus we have @xmath451 and @xmath452 .
changing variables in inequalities ( [ 2ndineq ] ) and ( [ 3rdineq ] ) as @xmath453 , @xmath454 , and @xmath455 ^ 2\rightarrow [ n(\theta\!=\!\pi)]^2 $ ] , we obtain @xmath456 ^ 2c}\bigl(1+\frac{a}{r } \bigr)\frac{a}{c}+\bigl(1+\frac{a}{r } \bigr)^5 \frac{\nu^2}{\sigma^2}-\frac{a}{c}+\bigl(1+\frac{a}{r } \bigr)>0~ , \label{2ndineqpi } \\ & & -\bigl\ { \frac{a}{c}- \bigl(1+\frac{a}{r } \bigr ) + \frac{mga}{[n(\theta=\pi)]^2c } \bigl(1+\frac{a}{r } \bigr)^2\bigr\}>0~ , \label{3rdineqpi}\end{aligned}\ ] ] for the stability for the inverted state . from the above two inequalities ,
we see : * in the case @xmath457 , i.e. , for the tippe top of group iii , the inequality ( [ 3rdineqpi ] ) is never satisfied . therefore , the inverted states ( @xmath213 ) of group iii are always unstable . * in the case @xmath458 , i.e. , for the tippe top of group i or ii , the inequality ( [ 3rdineqpi ] ) is satisfied if @xmath459 ^ 2>\frac{mga}{c\ { ( 1+\frac{a}{r } ) -\frac{a}{c}\}}\bigl(1 + \frac{a}{r } \bigr)^2 , \label{d7}\ ] ] which is the same requirement given in ( [ pistability3 ] ) for the stability of the tippe top of group i or ii at @xmath213 .
the inequality ( [ 2ndineqpi ] ) is automatically satisfied when both @xmath458 and inequality ( [ 3rdineqpi ] ) hold . actually ,
bmr derived also the stability criteria for the inverted state , taking @xmath460 , which are given by the three inequalities in ( bmr-(5.4))\!+\ ! \nu^2(1\!-\ ! e^\star)^7\!+\ ! ( 1\!-\ ! e^{\star})^3\mu e^{\star}fr^{-1 } ( 1\!-\!\mu { e^{\star } } ^2)>0 $ ] .
the error is traced back to the missing factor of @xmath461 in the expression of @xmath462 in bmr-(4.2 ) . ] . of course
, we can use them to obtain the stability conditions for the inverted state .
taking now @xmath463 and @xmath464 in the second and third inequality in ( bmr-(5.4 ) ) , we reach the same conclusions , ( bi ) and ( bii ) .
bmr discussed in ref .
@xcite about the heteroclinic connection between the non - inverted and inverted states of the tippe top .
they used an energy - momentum argument to determine the asymptotic states of the tippe top and obtained the explicit criteria for the existence of a heteroclinic connection , which are given in theorem 6.2 and the appendix of ref .
@xcite . in terms of the classification criteria and conditions obtained in this paper
, the statement in bmr on the existence of a heteroclinic connection can be restated as follows : ( i ) a tippe top must belong to group ii in order to have a heteroclinic connection .
( ii ) further more , the initial spin @xmath289 should be larger than @xmath232 ( eq.([zerostability4 ] ) ) and @xmath309 ( eq.([critical1 ] ) ) so that a tippe top becomes unstable at @xmath150 and reaches the inverted position .
the requirements @xmath465 and @xmath466 , respectively , correspond to the criteria @xmath467 and @xmath468 in theorem 6.2 in bmr .
h. k. moffatt , y. shimomura and m. branicki , dynamics of an axisymmetric body spinning on a horizontal surface .
part i : stability and the gyroscopic approximation , " proc .
a * 460 * , 3643 - 3672 ( 2004 ) . | we reexamine a very classical problem , the spinning behavior of the tippe top on a horizontal table .
the analysis is made for an eccentric sphere version of the tippe top , assuming a modified coulomb law for the sliding friction , which is a continuous function of the slip velocity @xmath0 at the point of contact and vanishes at @xmath1 .
we study the relevance of the gyroscopic balance condition ( gbc ) , which was discovered to hold for a rapidly spinning hard - boiled egg by moffatt and shimomura , to the inversion phenomenon of the tippe top .
we introduce a variable @xmath2 so that @xmath3 corresponds to the gbc and analyze the behavior of @xmath2 .
contrary to the case of the spinning egg , the gbc for the tippe top is not fulfilled initially .
but we find from simulation that for those tippe tops which will turn over , the gbc will soon be satisfied approximately .
it is shown that the gbc and the geometry lead to the classification of tippe tops into three groups : the tippe tops of group i never flip over however large a spin they are given .
those of group ii show a complete inversion and the tippe tops of group iii tend to turn over up to a certain inclination angle @xmath4 such that @xmath5 , when they are spun sufficiently rapidly .
there exist three steady states for the spinning motion of the tippe top . giving a new criterion for stability
, we examine the stability of these states in terms of the initial spin velocity @xmath6 . and we obtain a critical value @xmath7 of the initial spin which is required for the tippe top of group ii to flip over up to the completely inverted position .
( 5,2)(-270,-520 ) ( 2.3,35)ynu - hepth-05 - 102 ( 2.3,20)july 2005 |
big bang cosmology predicts the existence of a background gas of free photons and neutrinos . the measured cosmic microwave background ( cmb ) radiation supports the applicability of standard cosmology back to photon decoupling which occured approximately one hundred thousand years after the big bang .
the relic neutrinos , on the other hand , have decoupled when the universe had a temperature of one mev and an age of just one second .
thus , a measurement of the relic neutrinos , with a predicted average number density of @xmath3 per light ( @xmath4 mev ) neutrino species @xmath5 , would provide a new window to the early universe .
their predicted number density is comparable to the one of the microwave photons .
however , since neutrinos interact only weakly , the relic neutrinos have not yet been detected directly in laboratory experiments @xcite .
illustration of a z - burst resulting from the resonant annihilation of an ultrahigh energy cosmic neutrino on a relic ( anti-)neutrino .
, width=188 ] recently , an indirect detection possibility for relic neutrinos has been discussed @xcite .
it is based on so - called z - bursts resulting from the resonant annihilation of ultrahigh energy cosmic neutrinos ( uhec@xmath6s ) with relic neutrinos into @xmath7 bosons @xcite ( cf .
[ illu ] ) . on resonance ,
the corresponding cross section is enhanced by several orders of magnitudes .
if neutrinos have non - vanishing masses @xmath8 for which there is rather convincing evidence in view of the apparent observation of neutrino oscillations @xcite the respective resonance energies , in the rest system of the relic neutrinos , correspond to @xmath9 with @xmath10 denoting the mass of the z boson . these resonance energies are , for neutrino masses of @xmath11 ev , remarkably close to the energies of the highest energy cosmic rays observed at earth by collaborations such as agasa @xcite , fly s eye @xcite , haverah park @xcite , hires @xcite , and yakutsk @xcite ( for a review , see ref .
indeed , it was argued @xcite that the ultrahigh energy cosmic rays ( uhecrs ) above the predicted greisen - zatsepin - kuzmin ( gzk ) cutoff @xcite around @xmath12 ev are mainly protons ( and , maybe , photons ) from z decay . in this way ,
one possibly also solves one of the outstanding problems of ultrahigh energy cosmic ray physics @xcite , namely the apparent observation of cosmic rays with energies above the gzk cutoff , in an elegant and economical way without invoking new physics beyond the standard model , except for neutrino masses .
the gzk puzzle hinges on the fact that nucleons with super - gzk energies have a short attenuation length of about @xmath13 mpc , due to inelastic interactions with the cosmic microwave background , while plausible astrophysical sources for those energetic particles are much farther away @xcite .
ultrahigh energy neutrinos produced at cosmological distances , on the other hand , can reach the gzk zone unattenuated and their resonant annihilation on the relic neutrinos could just result in the observed cosmic rays of the highest energies .
moreover , at present this annihilation process may be the only way to detect the relic neutrino background , a basic ingredient of our cosmological picture .
the z - burst hypothesis for the ultrahigh energy cosmic rays was discussed in many papers @xcite .
the required uhec@xmath6 fluxes were estimated in ref .
@xcite for different spectral indices . in ref .
@xcite , particle spectra were determined numerically for case studies which supported the z - burst scenario .
the effect of possible lepton asymmetries was studied in ref .
@xcite . in ref .
@xcite , the analysis of the z - burst mechanism was advocated as one of the few possibilities for an absolute neutrino mass determination . in the present paper
, we present the details of ( and extend ) our recent quantitative investigation of the z - burst scenario @xcite ( see also refs .
@xcite ) , where we have determined the required mass of the heaviest relic neutrino as well as the necessary ultrahigh energy cosmic neutrino flux via a maximum likelihood analysis .
but before we start this enterprise , we shall briefly review the current knowledge of the absolute scale of neutrino masses .
neutrinos almost certainly have non - vanishing masses and mix .
this follows from the apparent observation of neutrino oscillations whose evidence is compelling for atmospheric neutrinos @xcite , strong for solar neutrinos @xcite , and so - far unconfirmed for neutrinos produced in the laboratory and studied by the lsnd collaboration and others @xcite .
however , neutrino oscillations are sensitive only to the mass ( squared ) splittings @xmath14 , not to the individual masses , @xmath15 , themselves . only a lower bound on the mass of the heaviest neutrino can be derived from these observations , e.g. @xmath16 from the atmospheric mass splitting in a three neutrino flavour scenario and @xmath17 from the still allowed value of the lsnd mass splitting in a four flavour scenario ,
respectively ( for a recent compilation of global analyses on mass splitting and mixing parameters , see for example refs .
@xcite and references therein ) . for an investigation of the absolute scale of neutrino masses
one has to exploit different types of experiments , such as the search for mass imprints in the endpoint spectrum of tritium beta ( @xmath18 ) decay or the search for neutrinoless double beta ( @xmath19 ) decay .
at present , these direct kinematical measurements of neutrino masses provide only upper limits , e.g. @xmath20 with @xmath21 being the leptonic mixing matrix , for the effective mass measured in tritium @xmath18 decay @xcite , and @xmath22 for the majorana neutrino mass parameter @xcite appearing in @xmath19 decay . the recently reported evidence for neutrinoless double beta decay and correspondingly deduced parameter range @xcite @xmath23 has been seriously challenged by refs .
combining the experimental constraints from oscillations and from tritium @xmath18 decay , one infers upper bounds on the mass of the heaviest neutrino @xcite , @xmath24 in a three flavour , and @xmath25 in a four flavour scenario , respectively .
for further recent investigations of the relationship between oscillation phenomena , @xmath18 decay and @xmath19 decay , in particular with respect to the neutrino mass spectrum , one may also consult refs .
@xcite and references cited therein .
further information on the absolute scale of neutrino masses can be obtained through cosmological and astrophysical considerations .
neutrinos in the @xmath26 ev mass range have cosmological implications since they represent a non - negligible part of dark matter .
this gives the opportunity to put upper limits on neutrino masses from cosmology @xcite .
analyses of galaxy clustering , including recent cmb measurements and other cosmological constraints , give an upper bound @xmath27 on the sum of the neutrino masses @xcite . from the spread of arrival times of neutrinos from supernova sn 1987a , coupled with the measured neutrino energies ,
a time - of - flight limit of @xmath28 ev can be derived @xcite , which , however , is not competitive with the direct limit ( [ lim_beta ] ) .
it is extremely welcome that the z - burst scenario opens a further and timely window to the absolute neutrino mass scale , since the opportunities to determine this crucial quantity are rare at present .
in addition , its verification would give us an indirect detection of the so elusive relic neutrinos and , finally , offer an explanation of the origin of the highest energy cosmic rays .
the organization of the present paper is as follows . in section [
sect : spectra ] we describe our determination of the proton and photon spectra at earth , which originate from z - bursts taking place mainly at extragalactic distances .
we discuss in detail the main ingredients in the predictions of the spectra , such as the details of z production and hadronic decay , the propagation of nucleons and photons through the diffuse extragalactic photon background , the diffuse flux of ultrahigh energy cosmic neutrinos , and the relic neutrino number density , as well as their anticipated uncertainties .
the comparison of the z - burst spectra with the observed ultrahigh energy cosmic ray spectrum is presented in section [ sect : determination ] .
the required absolute neutrino mass , as well as the necessary ultrahigh energy cosmic neutrino flux , are determined via a maximum likelihood analysis for various assumptions about the nature of the background of ordinary cosmic rays from unresolved astrophysical sources and variations of the diffuse extragalactic photon background notably in the radio band .
on the basis of these studies we find that the z - burst determinations of the mass of the heaviest neutrino as well as of the neutrino flux are fairly robust .
we find a required neutrino mass range of @xmath0 ev @xmath1 ev at the 68% confidence level , if the background of ordinary cosmic rays is of extragalactic origin .
this range narrows down considerably if a particular universal radio background is assumed , e.g. to @xmath0 ev @xmath2 ev for a large one . in section [
sect : discussion ] we discuss the implications of our findings for future laboratory studies such as the tritium beta decay and neutrinoless double beta decay , as well as for astrophysical and cosmological neutrino investigations .
this section contains also our conclusions .
our comparison of the z - burst scenario with the observed uhecr spectrum proceeds as follows .
first , we determine the probability of z production as a function of the distance from earth . secondly , we exploit collider experiments to derive the energy distribution of the produced protons and photons in the laboratory ( lab ) system .
thirdly , we consider the propagation of the protons and photons , i.e. we determine their energy losses due to pion and/or @xmath29 production through scattering on the diffuse extragalactic background photons and due to their redshift .
the last step is the comparison of the predicted and the observed spectra and the extraction of the required mass of the heaviest relic neutrino and of the necessary uhec@xmath6 flux .
our prediction of the differential proton flux , i.e. the number of protons arriving at earth with energy @xmath30 per units of energy ( @xmath30 ) , of area ( @xmath31 ) , of time ( @xmath32 ) and of solid angle ( @xmath33 ) , @xmath34 from z - bursts can be summarized as @xmath35 \nonumber & & \times \left [ \int\limits_0^\infty { \rm d } e_{\nu_i}\,f_{\nu_i}(e_{\nu_i},r)\ , n_{\bar\nu_i}(r ) \right .
\\[1ex]\nonumber & & \left .
\hspace{6ex } + \int\limits_0^\infty { \rm d } e_{\bar\nu_i}\ , f_{\bar\nu_i}(e_{\bar\nu_i},r)\ , n_{\nu_i}(r ) \right ] \\[1ex ] \nonumber & & \times \sigma_{\nu_i\bar\nu_i}(s)\ , { \rm br}(z\to { \rm hadrons})\ , \frac{{\rm d } n_{p+n}}{{\rm d } e_p } \\[1ex ] \nonumber & & \times ( -)\frac{\partial}{\partial e } p_p(r , e_p;e ) \,,\end{aligned}\ ] ] with the following important building blocks : the uhec@xmath6 fluxes @xmath36 at the energies @xmath37 ( @xmath38 ) and at the distance @xmath39 of z production to earth , the number density @xmath40 of the relic neutrinos , the z production cross section @xmath41 at centre - of - mass ( cm ) energy squared @xmath42 , the branching ratio @xmath43 , the energy distribution @xmath44 of the produced protons ( and neutrons ) with energy @xmath45 , and the probability @xmath46 that a proton created at a distance @xmath39 with energy @xmath45 arrives at earth above the threshold energy @xmath30 .
a similar expression as eq . ( [ p - flux ] ) holds for the differential photon flux from z - bursts , @xmath47 \nonumber & & \times \left [ \int\limits_0^\infty { \rm d } e_{\nu_i}\,f_{\nu_i}(e_{\nu_i},r)\ , n_{\bar\nu_i}(r ) \right .
\\[1ex ] \nonumber & & \left .
\hspace{6ex } + \int\limits_0^\infty { \rm d } e_{\bar\nu_i}\,f_{\bar\nu_i}(e_{\bar\nu_i},r)\ , n_{\nu_i}(r ) \right ] \\[1ex ] \nonumber & & \times \sigma_{\nu_i\bar\nu_i } ( s)\ , { \rm br}(z\to { \rm hadrons})\ , \frac{{\rm d } n_\gamma}{{\rm d } e_\gamma } \\[1ex ] \nonumber & & \times ( -)\frac{\partial}{\partial e } p_\gamma ( r , e_\gamma ; e ) \,.\end{aligned}\ ] ] here , the photon propagation function @xmath48 gives the expected number of detected photons above the threshold energy @xmath30 if one photon started from a distance of @xmath39 with energy @xmath49 .
note , that the @xmath50 function has a different interpretation than the @xmath51-function of protons .
this arises from the fact that the number of photons in distinction to the number of protons is not conserved during their propagation .
the building blocks related to z - production and decay
@xmath52 , the hadronic branching ratio , and the momentum distributions @xmath53 are very well , and the propagation functions @xmath54 are fairly well determined , whereas the first two ingredients , the flux of uhec@xmath6s , @xmath36 , and the radial distribution of the relic neutrino number density @xmath40 , are much less accurately known . in the following we shall discuss all these ingredients in detail . at lep and slc millions of z bosons were produced and their decays analyzed with extreme high accuracy . due to the large statistics , the uncertainties of our analysis related to z decay turned out to be negligible .
let us start with a review of the standard model neutrino annihilation cross section .
the s - channel z - exchange annihilation cross section into any fermion anti - fermion ( @xmath55 ) pair is given by ( see e.g. refs .
@xcite ) @xmath56 where @xmath57 is the cm energy squared , @xmath58 is the total width of the z boson , and @xmath59 is the effective number of annihilation channels , @xmath60 \nonumber & & \times \ , \frac{2}{3}\,n_f \left ( 1 - 8\ , t_{3\,f}\ , q_f\ , \sin^2\theta_w + 8\,q_f^2\sin^4\theta_w\right ) \,.\end{aligned}\ ] ] here the sum is over all fermions with @xmath61 , with charge @xmath62 ( in units of the proton charge ) , isospin @xmath63 ( @xmath64 for @xmath65 and neutrinos ; @xmath66 for @xmath67 and negatively charged leptons ) , and @xmath68 for leptons ( quarks ( @xmath69 ) ) . with @xmath70 for the @xmath71 of the effective weinberg angle and @xmath72 gev@xmath73 for the fermi coupling constant @xcite , formula ( [ z - res - cs - roulet ] )
gives , at the @xmath7-mass , @xmath74 \label{cs - tot - zpeak } \sigma ( \nu_i \bar{\nu}_i \to z^\ast \to { \rm all}\ f\bar f)\mid_{s = m_z^2 } & = & 455.6\ { \rm nb}\,,\end{aligned}\ ] ] with a branching ratio @xmath75 , in good agreement with the experimental result @xcite , @xmath76 later we shall exploit the following simplification which arises due to the fact that the cross section ( [ z - res - cs - roulet ] ) is sharply peaked at the resonance cm energy squared @xmath77 .
correspondingly , it acts essentially like a @xmath78-function in the integration over the energies @xmath37 of the incident neutrinos in eqs .
( [ p - flux ] ) and ( [ ph - flux ] ) , and we can assume that the uhec@xmath6 fluxes are constant in the relevant energy region .
thus , introducing the energy - averaged annihilation cross section @xcite , @xmath79 which is the effective cross section for all neutrinos within @xmath80 of their peak annihilation energy , we can write @xmath81\nonumber & & \hspace{8ex } \simeq e_{\nu_i}^{\rm res}\,f_{\nu_i}(e_{\nu_i}^{\rm res})\,\langle \sigma_{\rm ann}\rangle \,.\end{aligned}\ ] ] in view of the expected rapid decrease of the uhec@xmath6 flux at increasing energies ( cf .
section [ fluxes ] ) , we neglect @xmath32-channel w- and z - exchange annihilation processes . on resonance ,
the @xmath57-channel z - exchange processes completely overwhelm them . fairly above the resonant energies , on the other hand , they eventually dominate but are probably unobservable due to the lack of an appreciable uhec@xmath6 flux .
next , we turn to the energy distribution of the protons and photons in z decay .
we combined existing published and some improved unpublished data on the momentum distribution , @xmath82 of protons ( @xmath83 ) ( plus antiprotons ( @xmath84 ) ) in z decays @xcite , see fig .
[ prot - mom - dist ] ( top ) .
the experimental data , ranging down to @xmath85 , were combined with the predictions from the modified leading logarithmic approximation ( mlla ) @xcite at low @xmath86 .
the @xmath87 multiplicity is @xmath88 in the hadronic channel @xcite .
momentum distributions in hadronic z decays .
_ top : _ combined data from refs .
@xcite on proton ( plus antiproton ) momentum distribution ( solid ) , normalized to @xmath89 , and charged pion momentum distributions ( dotted ) , normalized to @xmath90 .
_ bottom : _ distribution of protons ( `` p '' ; solid ) , photons ( `` @xmath91 '' ; dotted ) and electrons ( `` e '' ; dashed ) in the lab system , in which the target relic neutrino is at rest .
, title="fig:",width=326 ] momentum distributions in hadronic z decays . _
top : _ combined data from refs .
@xcite on proton ( plus antiproton ) momentum distribution ( solid ) , normalized to @xmath89 , and charged pion momentum distributions ( dotted ) , normalized to @xmath90 . _
bottom : _ distribution of protons ( `` p '' ; solid ) , photons ( `` @xmath91 '' ; dotted ) and electrons ( `` e '' ; dashed ) in the lab system , in which the target relic neutrino is at rest .
, title="fig:",width=326 ] in the cm system of the z production the angular distribution of the hadrons is determined by the spin @xmath64 of the primary quarks and thus proportional to @xmath92 ( here @xmath93 is the angle between the incoming neutrinos and the outgoing hadrons ( cf .
@xcite ) ) .
the energy distribution of the produced protons with energy @xmath45 entering the z - burst spectrum ( [ p - flux ] ) , @xmath94 with @xmath95 , is finally obtained after a lorentz transformation from the cm system to the lab system , @xmath96 where @xmath97 is the proton mass .
the first line comes from the angular distribution , the second line is the jacobian and the third one is the momentum distribution at the inverted momentum .
the scaled energy distribution @xmath98 , as defined in eq .
( [ q - def ] ) and given by eq .
( [ q - dist ] ) , is displayed in fig .
[ prot - mom - dist ] ( bottom ) .
neutrons produced in z decays will decay and end up as uhecr protons .
they are taken into account according to @xmath99 where the neutron ( @xmath100 ) ( plus antineutron ( @xmath101 ) ) multiplicity , @xmath102 , is @xmath103 smaller than the proton s @xcite .
photons are produced in hadronic z decays via fragmentation into neutral pions , @xmath104 ( cf .
[ illu ] ) . the corresponding scaled energy distribution in the lab system , defined analogously to eq . ( [ q - def ] )
, reads @xmath105 where the scaled energy distribution @xmath106 , with @xmath107 , is given by eq .
( [ q - dist ] ) , with @xmath108 and @xmath109 , the momentum distribution of pions in hadronic z decay . for the latter distribution ,
we take the measured one of charged pions @xmath110 from hadronic z decay @xcite ( cf .
[ prot - mom - dist ] ( top ) ) , normalized such that @xmath111 @xcite .
note , that the distributions @xmath112 , @xmath113 , presented in fig .
[ prot - mom - dist ] ( bottom ) , compare favorably with the ones presented in ref .
@xcite based on the event generator pythia @xcite .
electrons ( and positrons ) from hadronic z decay are also relevant for the development of electromagnetic cascades .
they stem from decays of secondary charged pions , @xmath114 ( cf .
fig . [ illu ] ) , and their scaled energy distribution in terms of @xmath115 reads @xmath116 \nonumber & & \times { \mathcal q}_{\pi^\pm } \left ( \frac{2\ , y}{xw+\sqrt{x^2+(2\,m_e / m_\pi)^2 } } \right ) { \mathcal p}_e ( x ) \,,\end{aligned}\ ] ] where @xmath117 is the momentum distribution of the electrons in the rest system of the charged pion .
the energy distribution @xmath118 is also displayed in fig .
[ prot - mom - dist ] ( bottom ) .
the cosmic microwave background is known to a high accuracy .
it plays the key role in the determination of the probability @xmath46 that a proton created at a distance @xmath39 with energy @xmath45 arrives at earth above the threshold energy @xmath30 , suggested in ref .
@xcite and determined for a wide range of parameters in ref .
the propagation function @xmath119 takes into account the fact that protons of extragalactic ( eg ) origin and energies above @xmath120 ev lose a large fraction of their energy due to pion and @xmath29 production through scattering on the cmb and due to their redshift @xcite . in the present analysis
we have included new results for @xmath46 which include now variations in the cosmological parameters , in extension to the already published form @xcite exploited in ref .
@xcite . in our analysis
we go , according to @xmath121 out to distances @xmath122 ( cf .
( [ p - flux ] ) ) corresponding to redshift @xmath123 ( cf .
we use the expression @xmath124\ ] ] for the relation of the hubble expansion rate at redshift @xmath125 to the present one .
uncertainties of the latter , @xmath126 100 km / s / mpc , with @xmath127 @xcite , are included . in eq .
( [ h - omega ] ) , @xmath128 and @xmath129 , with @xmath130 , are the present matter and vacuum energy densities in terms of the critical density . as default values
we choose @xmath131 and @xmath132 , as favored today .
our results turn out to be pretty insensitive to the precise values of the cosmological parameters .
_ top : _ the intensity spectrum of the diffuse extragalactic photon background at redshift @xmath133 from ref .
@xcite ( solid ) .
different estimates from ref .
@xcite of the universal radio background ( urb ) are also indicated : high urb ( long - dashed ) and moderate urb ( short - dashed ) .
_ bottom : _ photon energy attenuation length @xmath134 at @xmath133 corresponding to the photon background shown above @xcite ( solid ) .
variations of @xmath134 arising from different assumptions about the urb ( cf . ref .
@xcite and top ) are also indicated : high urb ( long - dashed ) and moderate urb ( short - dashed ) .
shown is furthermore the energy attenuation length for electrons due to synchrotron radiation ( long - dashed - short - dashed ) , for different magnitudes of the extragalactic magnetic fields .
, title="fig:",width=326 ] _ top : _ the intensity spectrum of the diffuse extragalactic photon background at redshift @xmath133 from ref .
@xcite ( solid ) .
different estimates from ref .
@xcite of the universal radio background ( urb ) are also indicated : high urb ( long - dashed ) and moderate urb ( short - dashed ) .
_ bottom : _ photon energy attenuation length @xmath134 at @xmath133 corresponding to the photon background shown above @xcite ( solid ) .
variations of @xmath134 arising from different assumptions about the urb ( cf . ref .
@xcite and top ) are also indicated : high urb ( long - dashed ) and moderate urb ( short - dashed ) .
shown is furthermore the energy attenuation length for electrons due to synchrotron radiation ( long - dashed - short - dashed ) , for different magnitudes of the extragalactic magnetic fields .
, title="fig:",width=326 ] the simulations needed for the computation of @xmath46 require large computer power .
we have used a farm of personal computers ( pcs ) consisting of 128 parallel @xmath135 ghz pentium 4 processors normally devoted to qcd lattice calculations @xcite and have exploited about 3 hours cpu time to determine @xmath46 in a range of @xmath136 mpc , @xmath137 ev @xmath138 ev , and @xmath137 ev @xmath139 ev , for each fixed value of the cosmological parameters . the simulation was carried out in small ( @xmath140 kpc ) steps in @xmath39 . for each step , the statistical energy losses due to pion/@xmath29 production and redshift are taken into account @xcite . in this connection , the advantage of our formulation of the z - burst spectrum ( cf .
( [ p - flux ] ) ) in terms of the probability @xmath46 becomes evident .
we have to determine the latter only once and for all . without the use of @xmath46
, we would have to perform a simulation for any variation of the input spectrum , notably for any change in the neutrino mass .
our maximum likelihood analysis , involving the neutrino mass as a free parameter , would thus require excessive computer power on the order of 300 hours on the above pc farm for fixed cosmological parameters .
since @xmath46 is of universal usage , we have decided to make the corresponding numerical data for the probability distribution @xmath141 available for the public via the world - wide - web url _ http://www.desy.de/~uhecr _ .
the determination of the photon propagation function @xmath48 , entering the photon flux prediction ( [ ph - flux ] ) , was done as follows . in distinction to the case of the proton propagation function , we used here the continuous energy loss ( cel ) approximation which largely simplifies the work and reduces the required computer resources we have estimated the necessary cpu time for a full simulation on the pc farm mentioned above to 1000 hours per fixed set of the cosmological parameters . in the cel approximation ,
the energy ( and number ) of the detected photons is a unique function of the initial energy and distance , and statistical fluctuations are neglected
. a full simulation of the photon propagation function will be the subject of a later work .
the processes that are taken into account are pair production on the diffuse extragalactic photon background ( cf .
[ ph_mfp ] ( top ) ) , double pair production and inverse compton scattering of the produced pairs .
we comment also on synchrotron radiation in a possible extragalactic magnetic field ( egmf ) . for the energy attenuation length of the photons due to these processes
, we exploited the values quoted in ref .
@xcite ( see also ref .
@xcite ) and the further ones presented in fig .
[ ph_mfp ] ( bottom ) which incorporate various assumptions about the poorely known universal radio ( urb ) ( from ref .
@xcite ) and infrared ( irb ) backgrounds
. we shall analyse later the dependence of the neutrino mass and other fit parameters on these variations .
note , that , in view of the recent urb estimates in ref .
@xcite , the ones presented in fig .
[ ph_mfp ] ( top ) ) , which are based on ref .
@xcite , can be referred to as `` minimal '' urb .
the computation of the photon propagation function @xmath48 was carried out in the following way .
the energy attenuation of photons in the cel approximation was calculated according to @xmath142 where @xmath143 is the energy attenuation length at redshift @xmath125 .
the number of photons was assumed to be constant at ultrahigh energies @xmath144 ev , due to the small inelasticities in this energy range .
below , it was increased in a way to maintain energy conservation ( except for the redshift contribution ) : @xmath145 the @xmath48 function was then obtained by integration of these equations . in the ultrahigh energy region which is most relevant for us since we perform our fit to the cosmic ray data there
the approximation described above gives the photon flux quite reliable , while at lower energies it yields an upper bound .
presently unknown ingredients in the evaluation of the z - burst spectra ( [ p - flux ] ) and ( [ ph - flux ] ) are the differential fluxes @xmath146 of ultrahigh energy cosmic neutrinos ( see e.g. refs .
@xcite for recent reviews ) .
present experimental upper limits on these fluxes are rather poor ( cf . fig .
[ flux_upp_lim ] and refs .
@xcite ) .
upper limits on differential neutrino fluxes in the ultrahigh energy regime . shown
are experimental upper limits on @xmath147 from fly s eye @xcite and on @xmath148 from the goldstone lunar ultrahigh energy neutrino experiment glue @xcite , as well as theoretical upper limits on @xmath149 from `` visible '' ( `` wb '' @xcite ) and `` hidden '' ( `` mpr ''
@xcite ) hadronic astrophysical sources .
, width=302 ] what are the theoretical expectations for diffuse uhec@xmath6 fluxes ? more or less guaranteed are the so - called cosmogenic neutrinos which are produced when ultrahigh energy cosmic protons scatter inelastically off the cosmic microwave background radiation @xcite in processes such as @xmath150 , where the produced pions subsequently decay @xcite .
these fluxes ( for recent estimates , see refs .
@xcite ) represent reasonable lower limits , but turn out to be insufficient for the z - burst scenario .
recently , theoretical upper limits on the ultrahigh energy cosmic neutrino flux have been given in refs .
@xcite . per construction ,
the upper limit from `` visible '' hadronic astrophysical sources , i.e. from those sources which are transparent to ultrahigh energy cosmic protons and neutrons , is of the order of the cosmogenic neutrino flux and shown in fig .
[ flux_upp_lim ] ( `` wb '' ; cf
. refs .
@xcite ) . also shown in this figure ( `` mpr '' ) is the much larger upper limit from `` hidden '' hadronic astrophysical sources , i.e. from those sources from which only photons and neutrinos can escape @xcite .
even larger fluxes at ultrahigh energies may arise if the hadronic astrophysical sources emit photons only in the sub - mev region thus evading the `` mpr '' bound in fig .
[ flux_upp_lim ] or if the neutrinos are produced via the decay of superheavy relic particles @xcite , for which also the fragmentation function of the decay is of major interest @xcite , or topological defects @xcite .
in this situation of insufficient knowledge , we take the following approach concerning the flux of ultrahigh energy cosmic neutrinos , @xmath36 . it is assumed to have the form @xmath151 where @xmath125 is the redshift and where @xmath152 characterizes the cosmological source evolution ( see also refs .
the flux at zero redshift , @xmath153 , is left open . for hadronic
astrophysical sources it is expected to fall off power - like , @xmath154 , @xmath155 , at high energies . due to this fact and
because of the strong resonance peaks in the @xmath156 annihilation cross section ( [ z - res - cs - roulet ] ) at the resonance energies ( [ eres ] ) , the z - burst rate will be only sensitive to the flux at the resonant energy of the heaviest neutrino .
of course , the latter may be nearly degenerate with the other neutrino mass eigenstates , @xmath157 , as it is the case for @xmath158 ev in a three flavour scenario .
correspondingly , our later fit to the uhecr data will be sensitive only to @xmath159 \,,\ ] ] where the sum extends over the number of mass eigenstates which are quasi - degenerate with the heaviest neutrino .
note , finally , that , independently of the production mechanism , neutrino oscillations result in a uniform @xmath146 mixture for the different mass eigenstates @xmath5 .
_ top : _ mass density fluctuation field @xmath78 along the supergalactic plane as obtained from peculiar velocity measurements @xcite .
shown are contours in intervals of @xmath160 , surface maps on a grid of spacing @xmath161 kms@xmath162 , corresponding to @xmath163 mpc , with the height proportional to @xmath78 , and contrast maps .
one recognizes some well - known structures in the nearby volume such as the great attractor at supergalactic coordinates ( sgx @xmath164 kms@xmath162 , sgy @xmath165 kms@xmath162 ) , the perseus - pisces complex ( sgx @xmath166 kms@xmath162 , sgy @xmath167 kms@xmath162 ) , and the large void ( sgx @xmath168 kms@xmath162 , sgy @xmath169 kms@xmath162 ) in between . _
bottom : _ mass density fluctuation field obtained from above data , averaged over all directions , for @xmath170 .
the overdensities at around 20 and 80 mpc reflect the great attractor and the perseus - pisces complex , respectively . ,
title="fig:",width=453 ] _ top : _ mass density fluctuation field @xmath78 along the supergalactic plane as obtained from peculiar velocity measurements @xcite . shown
are contours in intervals of @xmath160 , surface maps on a grid of spacing @xmath161 kms@xmath162 , corresponding to @xmath163 mpc , with the height proportional to @xmath78 , and contrast maps .
one recognizes some well - known structures in the nearby volume such as the great attractor at supergalactic coordinates ( sgx @xmath164 kms@xmath162 , sgy @xmath165 kms@xmath162 ) , the perseus - pisces complex ( sgx @xmath166 kms@xmath162 , sgy @xmath167 kms@xmath162 ) , and the large void ( sgx @xmath168 kms@xmath162 , sgy @xmath169 kms@xmath162 ) in between . _
bottom : _ mass density fluctuation field obtained from above data , averaged over all directions , for @xmath170 .
the overdensities at around 20 and 80 mpc reflect the great attractor and the perseus - pisces complex , respectively .
, title="fig:",width=326 ] the dependence of the relic neutrino number density @xmath171 on the distance @xmath39 is treated in the following way .
the question is whether there is remarkable clustering of the relic neutrinos within the local gzk zone of about 50 mpc .
it is known that the density distribution of relic neutrinos as hot dark matter follows the total mass distribution ; however , with less clustering @xcite .
in fact , for @xmath172 ev , one expects pretty much that the neutrino number density equals the big bang prediction ( [ standard_number_dens ] ) @xcite . to take above facts into account ,
the shape of the @xmath40 distribution is varied , for distances below 100 mpc , between the standard cosmological homogeneous case ( [ standard_number_dens ] ) and that of the total mass distribution obtained from peculiar velocity measurements @xcite ( cf .
[ dens - prof ] ( top ) ) .
these peculiar measurements suggest relative overdensities of at most a factor @xmath173 , depending on the grid spacing ( cf .
[ dens - prof ] ( bottom ) ) .
a relative overdensity @xmath174 in our neighbourhood , as it was assumed in earlier investigations of the z - burst hypothesis @xcite , seems unlikely in view of these data .
our quantitative results turned out to be rather insensitive to the variations of the overdensities within the considered range . for scales larger than 100 mpc
the relic neutrino density is taken according to the big bang cosmology prediction , @xmath175 @xmath176 .
possible uniform neutrino density enhancements due to eventual lepton asymmetries , as advocated in ref .
@xcite , are negligible in view of the recent , very stringent bounds on the neutrino degeneracies @xcite . in any case , such uniform enhancements change only the normalization of the z - burst spectra and have no effect on their shape
. therefore , they could possibly milder the required uhec@xmath6 flux , but the value of the neutrino mass inferred from the z - burst scenario is not affected by them .
the predicted spectra of protons and photons from z - bursts , ( [ p - flux ] ) and ( [ ph - flux ] ) , can now be compared with the observed uhecr spectrum ( cf .
[ fit_normal ] ) .
our analysis includes published uhecr data of agasa @xcite , fly s eye @xcite , haverah park @xcite , and hires @xcite , as well as unpublished one from the world wide web pages of the experiments on 17/03/01 ( for a review , see ref .
@xcite ) . due to normalization difficulties
we did not use the yakutsk @xcite results .
we shall take into account the fact that above @xmath12 ev less than @xmath177 of the cosmic rays can be photons at the @xmath178 confidence level ( c.l . )
@xcite ( see also refs .
@xcite ) . as usual , each logarithmic unit between @xmath179 and @xmath180
is divided into ten bins .
the predicted number of uhecr events in a bin is taken as @xmath181,\ ] ] where @xmath182 m@xmath183s@xmath184sr is the total exposure ( estimated from the highest energy events and the corresponding fluxes ) and where @xmath185 ev is the lower bound of the @xmath186 energy bin . the first term in eq .
( [ flux ] ) , @xmath187 , corresponds to the diffuse background of ordinary cosmic rays from unresolved astrophysical sources . below the gzk cutoff
, it should have the usual and experimentally observed power - law form @xcite .
the second term represents the sum of the proton and photon spectra , @xmath188 , eqs .
( [ p - flux ] ) and ( [ ph - flux ] ) , from z - bursts .
the separation of the flux into two terms ( one from the power - law background and one from the z - burst ) is physically well motivated .
the power - law part below the gzk cutoff is confirmed experimentally and , for extragalactic sources , it should suffer from the gzk effect . in the z - burst scenario ,
cosmic rays are coming from another independent source ( z - bursts ) , too .
what we observe is the sum of the two .
as the detailed fits in the next section will show , the flux from z - bursts is much smaller in the low energy region than the flux of the power - law background .
correspondingly , the low energy part of the spectrum ( between @xmath189 and @xmath190 ev ) has very little influence on the z - burst fit parameters , notably on the neutrino mass . the available uhecr data with their error bars and the best fits ( long - dashed ) from ordinary cosmic ray protons originating from our local neighbourhood ( `` halo background '' ; _ top _ ) and originating from diffuse extragalactic sources ( `` eg background '' ; _ bottom _ ) , respectively . for the latter case
, the bump at @xmath12 ev represents protons injected at high energies and accumulated just above the gzk cutoff due to their energy losses .
the predicted fall - off for energies above @xmath12 ev can be observed .
, title="fig:",width=298 ] the available uhecr data with their error bars and the best fits ( long - dashed ) from ordinary cosmic ray protons originating from our local neighbourhood ( `` halo background '' ; _ top _ ) and originating from diffuse extragalactic sources ( `` eg background '' ; _ bottom _ ) , respectively .
for the latter case , the bump at @xmath12 ev represents protons injected at high energies and accumulated just above the gzk cutoff due to their energy losses .
the predicted fall - off for energies above @xmath12 ev can be observed .
, title="fig:",width=298 ] we shall study several possibilities for the background term @xmath187 . the first one is based on the assumption that the diffuse background of ordinary cosmic rays even above the gzk cutoff at @xmath12 ev consists of protons which are produced in our neighborhood within our galactic halo or at least within the gzk zone of about 50 mpc . we shall refer to this possible cosmic ray background in the following as the `` halo background '' .
we should note , however , that there are no apparent suitable astrophysical sources within the gzk distance to which the highest energy events point @xcite a fact which is difficult to reconcile with the halo background model . for the halo background ,
no gzk attenuation is included .
the spectrum is assumed to have the usual power - law behavior which describes the data well for smaller energies @xcite ( cf .
[ fit_normal ] ( top ) ) , @xmath191\nonumber & & \hspace{6ex } \frac{a}{1\,{\rm ev}}\ , \left ( \frac{e}{1\,{\rm ev}}\right)^{-\beta } \hspace{3ex } ( { \rm halo\ bk'd } ) \,.\end{aligned}\ ] ] a similar background model , however with a fixed power - law index @xmath192 , as found by agasa to fit the data from @xmath193 ev ( cf . table v in ref .
@xcite ) , has been exploited in a recent z - burst simulation @xcite . in the second case
, we assume that the diffuse background of ordinary cosmic rays comes from protons which originate from uniformly distributed , extragalactic sources hence we name it `` extragalactic background '' . in view of the observed distribution of arrival directions of uhecrs
, this assumption seems to be phenomenologically more realistic than the halo background model .
though there are some peculiarly clustered events , the overall distribution at present statistics seems to be practically uniform @xcite .
the extragalactic background suffers of course from gzk attenuation . correspondingly , we take the above power - law ( [ pow - law - halo ] ) , @xmath194 , as an injection spectrum and take its modification due to the interactions with the cmb photons into account with the help of the proton propagation function @xmath46 , @xmath195\nonumber & & \hspace{3ex } \int_0^\infty { \rm d}e_p \int_0^{r_{\rm max } } { \rm d}r \,\left(1+z(r)\right)^3 \\[1ex ] \label{pow - law - eg } & & \hspace{5ex } \times\ , \frac{a}{1\,{\rm ev}}\ , \left ( \frac{e_p}{1\,{\rm ev}}\right)^{-\beta}\ , \\[1ex]\nonumber & & \hspace{5ex } \times\ , ( -)\frac{\partial p_p(r , e_p;e)}{\partial e } \hspace{3ex } ( { \rm eg\ bk'd})\,.\end{aligned}\ ] ] the predicted spectrum of the extragactic background protons shows an accumulation at around the gzk scale @xmath12 ev and a sharp drop beyond ( see fig .
[ fit_normal ] ( bottom ) ) .
as far as the z - burst spectra entering our prediction ( [ flux ] ) are concerned , we proceed as follows .
we shall mainly concentrate on the case where there is only one relevant neutrino mass scale @xmath196 , either because there are three neutrino types with nearly degenerate neutrino masses , @xmath157 , or there is one neutrino which is much heavier than the other ones such that the contribution of the latter to the cosmic ray spectrum can be neglected since the corresponding resonance energies are much larger and the uhec@xmath6 fluxes are expected to fall with increasing energy .
therefore , we fit only one neutrino mass parameter @xmath196 . in this case , the spectra ( [ p - flux ] ) and ( [ ph - flux ] ) for protons and photons from z - bursts can be written as @xmath197\nonumber & & b\ , \int_0^\infty { \rm d}e_i \int_0^{r_{\rm max } } { \rm d}r \,\left(1+z(r)\right)^{3+\alpha}\,\delta_n ( r ) \\[1ex]\label{fit - zburst - spectra } & & \frac{4\,m_\nu}{m_z^2}\ , { \mathcal q_i}\left ( y = \frac{4\,m_\nu\,e_i}{m_z^2 } \right ) \\[1ex]\nonumber & & \times\ , ( -)\frac{\partial p_i(r , e_i;e)}{\partial e } , \hspace{6ex } i = p,\gamma\,,\end{aligned}\ ] ] where @xmath152 is the cosmological evolution parameter , @xmath198 is the mass density fluctuation field ( cf . fig .
[ dens - prof ] ( bottom ) ) , normalized to one , and @xmath199 are the boosted momentum distributions from hadronic z decay , normalized to @xmath200 , for @xmath201 , and to @xmath202 , for @xmath203 .
we are left here with two fit parameters , the mass @xmath196 of the heaviest neutrino and the overall normalization @xmath204 , which may be expressed , on account of eqs .
( [ eresfres ] ) and ( [ fnures ] ) , in terms of the original quantities entering eqs . ( [ p - flux ] ) and ( [ ph - flux ] ) , as @xmath205 note , that the neglection of finite width effects , @xmath80 , in our implementation ( [ fit - zburst - spectra ] ) of the z - burst spectra is perfectly adequate in view of the relative errors @xmath206% which we will find later from our fits .
the expectation value for the number of events in a bin is given by eq .
( [ flux ] ) . to determine the most probable value for @xmath207 we use the maximum likelihood method and minimize @xcite the @xmath208 , @xmath209 \nonumber & & \sum_{\log\left(\frac{e_i}{{\rm ev}}\right)=18.5}^{\log\left(\frac{e_i}{\rm ev}\right)=26.0 } 2\left [ n(i)-n_{\rm
o}(i)+n_{\rm o}(i ) \ln\left ( n_{\rm o}(i)/n(i)\right ) \right],\end{aligned}\ ] ] where @xmath210 is the total number of observed events in the @xmath186 bin .
since the z - burst scenario results in a quite small flux for lower energies , we take the lower bound just below the `` ankle '' : @xmath211 ev .
our results are insensitive to the definition of the upper end ( the flux is extremely small there ) for which we choose @xmath212 .
the uncertainties of the measured energies are about 30% which is one bin . by means of a monte carlo analysis , we take these uncertainties into account and include the corresponding variations in our final error estimates . for comparison with recent work on the z - burst scenario @xcite ,
we have done also fits with no background component , @xmath213 in eq .
( [ flux ] ) , and a larger lower end , @xmath214 .
such a scenario , with @xmath215 ev , was advocated in ref .
@xcite as appropriate to explain all uhecr data above the gzk cutoff by the z - burst model and to attribute all uhecr data below the cutoff to ordinary astrophysical extragalactic sources .
qualitatively , our analysis can be understood as follows . in the z - burst scenario a small relic neutrino
mass needs a large incident neutrino energy @xmath216 ( [ eres ] ) in order to produce a z. large @xmath216 results in a large lorentz boost , thus large @xmath45 resp . @xmath49 . in this way
the _ shape _ of the detected energy ( @xmath30 ) spectrum determines the mass of the relic neutrino .
the sum of the necessary uhec@xmath6 fluxes @xmath217 , on the other hand , is determined by the over - all _ normalization _ @xmath204 .
our fitting procedure involves four parameters : @xmath218 and @xmath196 .
the minimum of the @xmath219 function is @xmath220 at @xmath221 which is the most probable value for the mass , whereas the 1@xmath222 ( 68% ) confidence interval for @xmath196 is determined by @xmath223 here @xmath224 , @xmath225 , @xmath226 are defined in such a way that the @xmath227 function is minimized in @xmath228 and @xmath204 , at fixed @xmath196 . the available uhecr data with their error bars and the best fits from z - bursts , for a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected ( @xmath229 ) .
_ top : _ best fit for the case of a halo background ( solid line ) .
the bump around @xmath12 ev is mainly due to the z - burst protons ( dash - dotted ) , whereas the almost horizontal contribution ( long - dashed ) is the first , power - law - like term of eq .
( [ flux ] ) .
_ middle : _ the case of an `` extragalactic '' uhecr background .
the first bump at @xmath12 ev represents protons produced at high energies and accumulated just above the gzk cutoff due to their energy losses .
the bump at @xmath230 ev is a remnant of the z - burst energy .
the long - dashed line shows the contribution of the power - law - like spectrum with the gzk effect included . _
bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=298 ] the available uhecr data with their error bars and the best fits from z - bursts , for a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected ( @xmath229 ) . _
top : _ best fit for the case of a halo background ( solid line ) .
the bump around @xmath12 ev is mainly due to the z - burst protons ( dash - dotted ) , whereas the almost horizontal contribution ( long - dashed ) is the first , power - law - like term of eq .
( [ flux ] ) . _
middle : _ the case of an `` extragalactic '' uhecr background .
the first bump at @xmath12 ev represents protons produced at high energies and accumulated just above the gzk cutoff due to their energy losses .
the bump at @xmath230 ev is a remnant of the z - burst energy .
the long - dashed line shows the contribution of the power - law - like spectrum with the gzk effect included . _
bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=298 ] the available uhecr data with their error bars and the best fits from z - bursts , for a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected ( @xmath229 ) .
_ top : _ best fit for the case of a halo background ( solid line ) .
the bump around @xmath12 ev is mainly due to the z - burst protons ( dash - dotted ) , whereas the almost horizontal contribution ( long - dashed ) is the first , power - law - like term of eq .
( [ flux ] ) .
_ middle : _ the case of an `` extragalactic '' uhecr background .
the first bump at @xmath12 ev represents protons produced at high energies and accumulated just above the gzk cutoff due to their energy losses .
the bump at @xmath230 ev is a remnant of the z - burst energy .
the long - dashed line shows the contribution of the power - law - like spectrum with the gzk effect included . _
bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=298 ] @xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath233 & @xmath234 & @xmath235 & @xmath236 & @xmath237 + @xmath238 & @xmath239 & @xmath240 & @xmath241 & @xmath242 & @xmath243 + @xmath244 & @xmath245 & @xmath246 & @xmath247 & @xmath248 & @xmath243 + + @xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath249 & @xmath250 & @xmath251 & @xmath252 & @xmath253 + @xmath238 & @xmath254 & @xmath255 & @xmath256 & @xmath257 & @xmath258 + @xmath244 & @xmath259 & @xmath260 & @xmath261 & @xmath262 & @xmath263 + + @xmath264 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath265 & @xmath266 & @xmath267 & @xmath268 & @xmath269 & @xmath268 + @xmath270 & @xmath271 & @xmath272 & @xmath268 & @xmath273 & @xmath268 + @xmath274 & @xmath275 & @xmath276 & @xmath268 & @xmath277 & @xmath268 + @xmath278 & @xmath279 & @xmath280 & @xmath268 & @xmath281 & @xmath268 + @xmath282 & @xmath283 & @xmath284 & @xmath268 & @xmath285 & @xmath268 + @xmath286 & @xmath287 & @xmath288 & @xmath268 & @xmath289 & @xmath268 + @xmath290 & @xmath291 & @xmath292 & @xmath268 & @xmath293 & @xmath268 + as already mentioned , presently there is no evidence that the observed highest energy cosmic rays are photons .
let us start therefore with the assumption ( cf . ref .
@xcite ) that the ultrahigh energy photons from z - bursts can be neglected in the fit in comparison to the protons .
this is certainly true for a sufficiently large universal radio background , e.g. on the level of the maximal one estimated in ref .
@xcite and/or for a sufficiently strong extragalactic magnetic field @xmath294
g. we shall refer to this scenario in the following as `` strong '' uhe@xmath91 attenuation .
our best fits to the observed data from this scenario ( cf .
table [ fit_noph ] ) can be seen in fig .
[ fit_nu ] , for evolution parameter @xmath295 .
we find a neutrino mass of @xmath296 ev for the case that the uhecr background protons are of halo type ( [ pow - law - halo ] ) , @xmath297 ev , if they are of extragalactic type ( [ pow - law - eg ] ) , and @xmath298 ev , if there are are no background protons above @xmath299 ev , respectively .
the first numbers are the 1@xmath222 , the numbers in the brackets are the 2@xmath222 errors .
this gives an absolute lower bound on the mass of the heaviest neutrino of @xmath300 ev at the 95% c.l .
, which is comparable to the one obtained from the atmospheric mass splitting in a three flavour scenario , eq .
( [ lim_low_atm ] ) . the available uhecr data with their error bars and the @xmath301 fits ( solid lines ) from z - bursts , for a halo ( _ top _ ) , an extragalactic ( _ middle _ ) , and no ( _ bottom _ ) uhecr background ( long - dashed lines ) , and a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected and only the z - burst protons ( dash - dotted lines ) have to be taken into account in the fit ( @xmath229 ) .
, title="fig:",width=325 ] the available uhecr data with their error bars and the @xmath301 fits ( solid lines ) from z - bursts , for a halo ( _ top _ ) , an extragalactic ( _ middle _ ) , and no ( _ bottom _ ) uhecr background ( long - dashed lines ) , and a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected and only the z - burst protons ( dash - dotted lines ) have to be taken into account in the fit ( @xmath229 ) . , title="fig:",width=325 ] the available uhecr data with their error bars and the @xmath301 fits ( solid lines ) from z - bursts , for a halo ( _ top _ ) , an extragalactic ( _ middle _ ) , and no ( _ bottom _ ) uhecr background ( long - dashed lines ) , and a strong uhe@xmath91 attenuation such that the z - burst photons can be neglected and only the z - burst protons ( dash - dotted lines ) have to be taken into account in the fit ( @xmath229 ) .
, title="fig:",width=325 ] the surprisingly small uncertainties are based on the @xmath227 analysis described in section [ gen ] .
the inclusion of the already mentioned 30% uncertainties in the observed energies by a monte carlo analysis increases the error bars by about 10% .
note , that the relative errors in the extragalactic and in the no background cases are of the same order .
this shows that the smallness of these errors does not originate from the low energy part of the background component .
the fits are rather good : for 21 non - vanishing bins and 4 fitted parameters they can be as low as @xmath302 and @xmath303 in the halo and the extralactic background case , respectively , whereas in the no background case , for @xmath304 ev , we have 9 bins with 2 fitted parameters and a @xmath305 , see table [ fit_noph ] . in the latter case , however , which was advocated strongly in ref .
@xcite , a remarkable dependence of the fitted mass on the value of @xmath306 is observed .
the @xmath307 fits are shown in fig .
[ fit_errors ] .
as it should be , the spread is small in the region where there are data , whereas it can be quite large in the presently unexplored ultrahigh energy regime .
finally , let us mention that , both in the `` halo '' as well as in the extragalactic background cases , the @xmath227 fits _ without _ a z - burst component ( cf .
[ fit_normal ] ) are far worse : we find @xmath308 in the former , and @xmath309 in the latter , for 21 non - vanishing bins and 2 fitted parameters .
this finding suggests that the power - law background terms alone can not describe the data .
in addition to the case of strong uhe@xmath91 attenuation , corresponding to a large universal radio background or a large extragalactic magnetic field , let us consider now the case of a `` minimal '' urb . here
, we assume a vanishing egmf and exploit , as in ref .
@xcite , a universal radio background on the level of the one from ref .
@xcite ( cf .
section [ prop ] and fig . [ ph_mfp ] ( top ) ) .
the corresponding fits , for our three background scenarios , can be seen in fig .
[ fit_nu_minurb ] and table [ fit_minurb ] .
note , that our best fits are still compatible with the already mentioned upper limits on the photon fraction of the observed ultrahigh energy cosmic rays @xcite .
we observe that the value of the neutrino mass found in both the halo background scenario , with @xmath310 ev , as well as in the extragalactic background scenario , with @xmath311 ev , are compatible , at about the 2 @xmath222 level , with the corresponding values found in the case of strong uhe@xmath91 attenuation . as before ,
in the `` no '' background scenario , we find a strong dependence of the fitted mass on the value of @xmath306 . the available uhecr data with their error bars and the best fits from z - bursts , for an energy attenuation of photons as in ref . @xcite and a vanishing egmf ( @xmath312 ) .
_ top : _ best fit for the case of a halo uhecr background ( solid line ) , corresponding to the sum of the background protons ( long - dashed ) , the z - burst protons ( dash - dotted ) and the z - burst photons ( short - dashed ) .
_ middle : _ the case of an extragalactic uhecr background .
_ bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=325 ] the available uhecr data with their error bars and the best fits from z - bursts , for an energy attenuation of photons as in ref .
@xcite and a vanishing egmf ( @xmath312 ) .
_ top : _ best fit for the case of a halo uhecr background ( solid line ) , corresponding to the sum of the background protons ( long - dashed ) , the z - burst protons ( dash - dotted ) and the z - burst photons ( short - dashed ) . _
middle : _ the case of an extragalactic uhecr background . _
bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=325 ] the available uhecr data with their error bars and the best fits from z - bursts , for an energy attenuation of photons as in ref .
@xcite and a vanishing egmf ( @xmath312 ) .
_ top : _ best fit for the case of a halo uhecr background ( solid line ) , corresponding to the sum of the background protons ( long - dashed ) , the z - burst protons ( dash - dotted ) and the z - burst photons ( short - dashed ) .
_ middle : _ the case of an extragalactic uhecr background .
_ bottom : _ the case of no uhecr background above @xmath231 .
, title="fig:",width=325 ] @xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath313 & @xmath314 & @xmath315 & @xmath316 & @xmath317 + @xmath238 & @xmath318 & @xmath319 & @xmath320 & @xmath321 & @xmath322 + @xmath244 & @xmath323 & @xmath324 & @xmath325 & @xmath326 & @xmath327 + + @xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath328 & @xmath329 & @xmath330 & @xmath331 & @xmath258 + @xmath238 & @xmath332 & @xmath333 & @xmath334 & @xmath335 & @xmath263 + @xmath244 & @xmath336 & @xmath337 & @xmath338 & @xmath339 & @xmath340 + + @xmath264 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath265 & @xmath341 & @xmath342 & @xmath268 & @xmath343 & @xmath268 + @xmath270 & @xmath344 & @xmath345 & @xmath268 & @xmath346 & @xmath268 + @xmath274 & @xmath347 & @xmath348 & @xmath268 & @xmath349 & @xmath268 + @xmath278 & @xmath350 & @xmath351 & @xmath268 & @xmath352 & @xmath268 + @xmath282 & @xmath353 & @xmath354 & @xmath268 & @xmath355 & @xmath268 + @xmath286 & @xmath356 & @xmath357 & @xmath268 & @xmath358 & @xmath268 + @xmath290 & @xmath359 & @xmath360 & @xmath268 & @xmath361 & @xmath268 + besides the no uhecr background scenario , the z - burst determination of the neutrino mass seems reasonably robust .
its robustness with regard to changes of presently unknown quantities within their anticipated variations is further illustrated in tables [ fit_modurb ] and [ fit_hubble ] and in fig .
[ mass_res ] . for a wide range of cosmological source evolution ( @xmath362 ) , hubble parameters @xmath363 , @xmath364 , @xmath365 , @xmath366 , for variations of the possible relic neutrino overdensity in our gzk zone within the limits discussed in section [ nunumb ] , and for different assumptions about the diffuse extragalactic photon background ,
the results remain within the above error bars .
the main uncertainties concerning the central values originate from the different assumptions about the background of ordinary cosmic rays . in the case that the ordinary cosmic rays above @xmath189 ev are protons and originate from a region within the gzk zone of about 50 mpc ( `` halo '' ) , the required mass of the heaviest neutrino seems to lie between @xmath367 ev @xmath368 ev at the 68% c.l .
( @xmath369 , cf .
below ) , if we take into account the variations between the minimal and moderate urb cases and the strong uhe@xmath91 attenuation case ( cf .
[ mass_res ] ) .
note , that a value of @xmath370 ev , as studied recently in ref .
@xcite in a simulation of the z - burst scenario in the context of a halo background model , seems to lie more than 3@xmath222 away from our best fit values .
the much more plausible assumption that the ordinary cosmic rays above @xmath189 are protons of extragalactic origin leads to a required neutrino mass of @xmath0 ev @xmath1 ev at the 68% c.l .
( @xmath369 ) ( cf .
[ mass_res ] ) .
@xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath371 & @xmath314 & @xmath372 & @xmath373 & @xmath374 + @xmath238 & @xmath375 & @xmath376 & @xmath377 & @xmath378 & @xmath379 + @xmath244 & @xmath380 & @xmath381 & @xmath382 & @xmath383 & @xmath384 + + @xmath152 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath232 & @xmath385 & @xmath386 & @xmath387 & @xmath388 & @xmath253 + @xmath238 & @xmath389 & @xmath390 & @xmath391 & @xmath392 & @xmath258 + @xmath244 & @xmath393 & @xmath394 & @xmath395 & @xmath396 & @xmath258 + + @xmath264 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath265 & @xmath397 & @xmath398 & @xmath268 & @xmath399 & @xmath268 + @xmath270 & @xmath400 & @xmath401 & @xmath268 & @xmath402 & @xmath268 + @xmath274 & @xmath403 & @xmath404 & @xmath268 & @xmath405 & @xmath268 + @xmath278 & @xmath406 & @xmath407 & @xmath268 & @xmath408 & @xmath268 + @xmath282 & @xmath409 & @xmath410 & @xmath268 & @xmath411 & @xmath268 + @xmath286 & @xmath412 & @xmath413 & @xmath268 & @xmath414 & @xmath268 + @xmath290 & @xmath415 & @xmath416 & @xmath268 & @xmath417 & @xmath268 + @xmath418 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath419 & @xmath420 & @xmath421 & @xmath422 & @xmath423 & @xmath424 + @xmath425 & @xmath426 & @xmath427 & @xmath428 & @xmath429 & @xmath430 + + @xmath418 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath419 & @xmath431 & @xmath432 & @xmath433 & @xmath434 & @xmath424 + @xmath425 & @xmath435 & @xmath436 & @xmath437 & @xmath438 & @xmath439 + + @xmath418 & @xmath196 [ ev ] & @xmath220 & @xmath31 & @xmath204 & @xmath18 + @xmath419 & @xmath440 & @xmath441 & @xmath442 & @xmath443 & @xmath444 + @xmath425 & @xmath445 & @xmath446 & @xmath447 & @xmath448 & @xmath430 + summary of the masses of the heaviest neutrino required in the z - burst scenario , with their 1@xmath222 ( solid ) and 2@xmath222 ( dotted ) error bars , for the case of an extragalactic and a halo background of ordinary cosmic rays and for various assumptions about the diffuse extragalactic photon background in the radio band ( @xmath229 ) . from left : strong @xmath91 attenuation , moderate and minimal urb . , width=326 ] we performed a monte carlo analysis studying higher statistics . in the near future ,
the pierre auger observatory @xcite will provide a ten times higher statistics , which reduces the error bars in the neutrino mass to about one third of their present values .
let us consider next in more detail the @xmath91 ray spectra from z - bursts , notably in the @xmath449 gev region . as illustrated in fig .
[ fit_egret ] , the egret measurements of the diffuse @xmath91 background in the energy range between 30 mev and 100 gev @xcite put non - trivial constraints on the evolution parameter @xmath152 . whereas different values of @xmath152 lead to very similar spectra in the ultrahigh energy region which determines the neutrino mass they are easily discriminated in @xmath450 tev photons .
only evolution parameters @xmath451 seem to be compatible with the egret measurements ( cf .
[ fit_egret ] ) , quite independently of different assumptions about the urb .
these numerical findings are in fairly good agreement with other recent simulations @xcite .
the necessary uhec@xmath6 flux at the resonant energy @xmath216 is obtained via eq .
( [ b - uhecnu - fluxes ] ) from the fitted overall normalization @xmath204 , @xmath452 where @xmath453 mpc is the present size of the universe and where , on account of eqs .
( [ z - h ] ) and ( [ h - omega ] ) , the ratio of the maximally considered distance of z production , @xmath122 , and the size of the universe is given by @xmath454 this ratio is of order one ; it equals @xmath455 for @xmath456 , @xmath457 and @xmath458 .
the required fluxes are summarized in fig .
[ eflux ] , together with some existing upper limits and projected sensitivities of present , near future and future observational projects .
they appear to be well below present upper limits and are within the expected sensitivity of amanda @xcite , rice @xcite , and auger @xcite . clearly , these fluxes are higher than the ones advocated in ref .
@xcite based on local neutrino overdensities @xmath459 on scales of about @xmath460 mpc . however , since we also took into account a background from ordinary cosmic rays , our normalization of the z - burst component is different and correspondingly our fluxes are somewhat less than a factor of @xmath461 higher .
they are consistent with the ones found in refs .
@xcite . as far as power - law extrapolations of the fluxes , @xmath462 , below the resonance energy are concerned , we find , in agreement with refs .
@xcite , that indices @xmath463 are excluded by the fly s eye limits @xcite ( cf .
[ eflux ] ) . the available uhecr data with their error bars and the best fit from z - bursts , for various cosmological evolution parameters @xmath152 and an energy attenuation of photons as in fig . [ ph_mfp ] ( bottom ) exploiting a `` minimal '' urb ( @xmath464 ) . also shown is the diffuse @xmath91 background in the energy range between 30 mev and 100 gev as measured by egret @xcite ( solid ) . , width=326 ]
the required neutrino flux , for power - law extrapolations with index @xmath465 to lower energies , is larger than the theoretical upper limit from `` hidden '' hadronic astrophysical sources , from which only photons and neutrinos can escape ( cf . fig .
[ flux_upp_lim ] ( `` mpr '' ) ) . from an analysis of the assumptions behind this limit
@xcite one finds that one has to invoke hidden astrophysical sources whose photons somehow do nt show up in the diffuse @xmath91 ray background measured by egret @xcite . in general , astrophysical sources for the required uhe neutrinos should be distributed with @xmath466 , accelerate protons up to energies @xmath467 ev , be opaque to primary nucleons and emit secondary photons only in the sub - mev region .
it is an interesting question whether such challenging conditions can be realized in bl lac objects , a class of active galactic nuclei for which some evidence of zero or negative cosmological evolution has been found ( see ref .
@xcite and references therein ) and which were recently discussed as possible sources of the highest energy cosmic rays @xcite .
alternatively , one may invoke top - down scenarios @xcite for the sources of the highest energy cosmic neutrinos such as unstable superheavy relic particle decays @xcite or topological defect decays @xcite .
it should be stressed that , besides the neutrino mass , the uhec@xmath6 flux at the resonance energy is one of the most robust predictions of the z - burst scenario which can be verified or falsified in the near future . in this connection
it is worthwhile to recall ( cf .
section [ nunumb ] ) that the current limits on neutrino degeneracies @xcite do not allow a remarkable uniform enhancement due to lepton asymmetries which otherwise would be welcome for a relaxation of the huge flux requirement @xcite .
neutrino fluxes , @xmath468 , required by the z - burst hypothesis for the case of a halo and an extragalactic background of ordinary cosmic rays , respectively ( @xmath229 ) . shown
are the necessary fluxes obtained from the fit results of table [ fit_noph ] for the case of a strong uhe@xmath91 attenuation .
the horizontal errors indicate the 1@xmath222 ( solid ) and 2@xmath222 ( dotted ) uncertainties of the mass determination and the vertical errors include also the uncertainty of the hubble expansion rate . also shown
are upper limits from fly s eye @xcite on @xmath147 and the goldstone lunar ultrahigh energy neutrino experiment glue @xcite on @xmath469 , as well as projected sensitivities of amanda @xcite on @xmath149 and auger @xcite on @xmath147 .
the sensitiviy of rice is comparable to the one of auger @xcite.,width=326 ]
we have presented a detailed comparison of the predicted spectra from z - bursts resulting from the resonant annihilation of ultrahigh energy cosmic neutrinos with relic neutrinos with the observed ultrahigh energy cosmic ray spectrum , extending our earlier study @xcite .
the mass of the heaviest relic neutrino turned out to have to lie in the range @xmath367 ev @xmath368 ev on the 1 @xmath222 level , if the background of ordinary cosmic rays above @xmath189 ev consists of protons and originates from a region within a distance of about 50 mpc . in the phenomenologically most plausible case that the ordinary cosmic rays above @xmath189 are protons of extragalactic origin one is lead to a required neutrino mass in the range @xmath0 ev @xmath1 ev at the 68% c.l .. we have also analysed a case where there is no background of ordinary cosmic rays above , say , @xmath299 ev . here , we find @xmath470 ev @xmath471 ev .
the above neutrino mass ranges include variations in presently unknown quantities , like the amount of neutrino clustering , the universal radio background , and the extragalactic magnetic field , within their anticipated uncertainties . if it turns out that the highest energy cosmic rays are extragalactic protons rather than photons one has to invoke a large universal radio background and/or large extragalactic magnetic field to suppress the photons from z - bursts .
in this case and for our extragalactic background scenario , the neutrino mass range narrows down to @xmath0 ev @xmath2 ev , with a best fit value of @xmath472 ev .
it is remarkable , that the mass ranges required in the z - burst scenario coincide nearly perfectly with the present knowledge about the mass of the heaviest neutrino from oscillations and tritium @xmath18 decay from eqs .
( [ lim_low_atm ] ) and ( [ lim_comb_osc_beta ] ) , @xmath473 ev , in a three flavour or , from eqs .
( [ lim_low_lsnd ] ) and ( [ lim_comb_osc_beta_lsnd ] ) , @xmath474 ev , in a four flavour scenario .
our values , in the extragalactic background scenario , are still compatible with a hierarchical neutrino mass spectrum , with the largest mass suggested by the atmospheric neutrino oscillation , @xmath475 ev . however , our best fit values point more into the direction of a quasi - degenerate scenario , @xmath476 . it is amusing to observe that the recently reported evidence for neutrinoless double beta decay , with @xmath477 ev , if true , would point also into the same direction @xcite .
the above neutrino masses are in a range which can be explored by near - future laboratory experiments , like the katrin tritium @xmath18 decay experiment @xcite , with a projected sensitivity of @xmath478 ev within @xmath479 years , or the @xmath19 decay experiments nemo-3 @xcite and cuore @xcite , which aim at @xmath480 ev , and their next - generation follow - up projects genius @xcite and exo @xcite , with @xmath481 ev .
our mass range also compares favourably with the expected sensitivity @xmath482 ev of near - future global cosmological analyses involving also new cmb measurements @xcite , as well as @xmath483 ev expected from neutrino observations from supernovae explosions @xcite . from our fits
, we could also determine the necessary ultrahigh energy neutrino flux at the resonance energy .
its prediction , in the context of the z - burst scenario , turned out to be of similar robustness as the prediction of the mass of the heaviest neutrino .
it was found to be consistent with present upper limits and detectable in the near future by the already operating neutrino telescopes amanda and rice , and by the pierre auger observatory , presently under construction .
a search at these facilities is the most promising and timely step in testing the z - burst hypothesis .
one does not have to wait for future projects such as euso @xcite for this critical investigation .
the required neutrino fluxes are enormous .
if such tremendous fluxes of ultrahigh energy neutrinos are indeed found , one has to deal with the challenge to explain their origin .
it is fair to say , that at the moment no convincing astrophysical sources are known which meet the requirements for the z - burst hypothesis , i.e. which have no or a negative cosmological evolution , accelerate protons at least up to @xmath484 ev , are opaque to primary nucleons and emit secondary photons only in the sub - mev region .
ideas to create the required neutrino fluxes from the decays of superheavy particles or topological defects do not seem to be too appealing since they invoke further new physics beyond neutrino masses .
we thank s. barwick , o. biebel , s. bludman , w. buchmller , p. di bari , m. kachelriess , k. mannheim , h. meyer , w. ochs , k. petrovay , s. sarkar , f. schrempp , d. semikoz , g. sigl , and e. waxman for useful discussions .
we thank the opal collaboration for their unpublished results on hadronic z decays .
this work was partially supported by hungarian science foundation grants no .
otka - t34980/t29803/t22929/m28413/omfb-1548/om - mu-708 .
l. stodolsky , phys .
* 34 * , 110 ( 1975 ) [ erratum - ibid . *
34 * , 508 ( 1975 ) ] .
f. smith and j. d. lewin , phys .
b * 127 * , 185 ( 1983 ) .
b. f. shvartsman , v. b. braginsky , s. s. gershtein , y. b. zeldovich and m. y. khlopov , jetp lett . * 36 * , 277 ( 1982 ) [ pisma zh .
eksp . teor .
fiz . * 36 * , 224 ( 1982 ) ] .
p. langacker , j. p. leveille and j. sheiman , phys .
d * 27 * , 1228 ( 1983 ) .
i. ferreras and i. wasserman , phys .
d * 52 * , 5459 ( 1995 ) . c. hagmann , astro - ph/9905258 .
g. duda , g. gelmini and s. nussinov , phys .
d * 64 * , 122001 ( 2001 ) .
d. fargion , b. mele and a. salis , astrophys . j. * 517 * , 725 ( 1999 ) . t. j. weiler , astropart .
phys . * 11 * , 303 ( 1999 ) .
t. weiler , phys .
. lett . * 49 * , 234 ( 1982 ) .
e. roulet , phys .
d * 47 * , 5247 ( 1993 ) .
s. yoshida , h. y. dai , c. c. jui and p. sommers , astrophys . j. * 479 * , 547 ( 1997 ) .
y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys . rev .
lett . * 81 * , 1562 ( 1998 ) .
y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys .
b * 433 * , 9 ( 1998 ) .
y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys .
b * 436 * , 33 ( 1998 ) .
s. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys . rev .
lett . * 85 * , 3999 ( 2000 ) .
m. ambrosio _ et al . _
[ macro collaboration ] , phys .
b * 517 * , 59 ( 2001 ) .
b. t. cleveland _ et al . _ , astrophys .
j. * 496 * , 505 ( 1998 ) .
w. hampel _ et al . _
[ gallex collaboration ] , phys .
b * 447 * , 127 ( 1999 ) .
j. n. abdurashitov _ et al .
_ [ sage collaboration ] , phys .
c * 60 * , 055801 ( 1999 ) .
s. fukuda _ et al .
_ [ superkamiokande collaboration ] , phys . rev .
* 86 * , 5651 ( 2001 ) .
s. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys .
lett . *
86 * , 5656 ( 2001 ) .
m. altmann _ et al .
_ [ gno collaboration ] , phys .
b * 490 * , 16 ( 2000 ) .
q. r. ahmad _ et al .
_ [ sno collaboration ] , phys . rev .
* 87 * , 071301 ( 2001 ) . c. athanassopoulos
[ lsnd collaboration ] , phys .
lett . *
75 * , 2650 ( 1995 ) . c. athanassopoulos
_ et al . _
[ lsnd collaboration ] , phys .
c * 54 * , 2685 ( 1996 ) . c. athanassopoulos
_ et al . _
[ lsnd collaboration ] , phys .
lett . * 77 * , 3082 ( 1996 ) . c. athanassopoulos _
[ lsnd collaboration ] , phys .
* 81 * , 1774 ( 1998 ) . c. athanassopoulos
_ et al . _
[ lsnd collaboration ] , phys .
c * 58 * , 2489 ( 1998 ) .
a. aguilar _ et al . _
[ lsnd collaboration ] , phys .
d * 64 * , 112007 ( 2001 ) .
y. declais _ et al .
_ , nucl .
b * 434 * , 503 ( 1995 ) .
m. apollonio _ et al .
_ [ chooz collaboration ] , phys .
b * 466 * , 415 ( 1999 ) .
f. boehm _ et al .
_ , phys .
* 84 * , 3764 ( 2000 ) .
k. eitel [ karmen collaboration ] , nucl .
suppl . * 91 * , 191 ( 2000 ) .
b. armbruster _ et al .
_ [ karmen collaboration ] , hep - ex/0203021 .
s. avvakumov _ et al .
_ , hep - ex/0203018 .
m. takeda _ et al .
_ , phys .
lett . * 81 * , 1163 ( 1998 ) .
d. j. bird _
[ hires collaboration ] , phys .
* 71 * , 3401 ( 1993 ) .
d. j. bird _
[ hires collaboration ] , astrophys .
j. * 424 * , 491 ( 1994 ) . d. j. bird _
et al . _ ,
astrophys .
j. * 441 * , 144 ( 1995 ) .
m. a. lawrence , r. j. reid and a. a. watson , j. phys .
g * 17 * , 733 ( 1991 ) .
m. ave , j. a. hinton , r. a. vazquez , a. a. watson and e. zas , phys .
lett . * 85 * , 2244 ( 2000 ) .
kieda00 d. kieda _ et al .
_ , in _ proceedings of the 26th international cosmic ray conference _ , salt lake , 1999 .
efimov91 n.n .
et al . _ , in _ proc .
of the astrophysical aspects of the most energetic cosmic rays _
( world scientific , singapore , 1991 ) .
m. nagano and a. a. watson , rev .
phys . * 72 * , 689 ( 2000 ) .
k. greisen , phys .
* 16 * , 748 ( 1966 ) .
g. t. zatsepin and v. a. kuzmin , jetp lett .
* 4 * , 78 ( 1966 ) [ pisma zh .
fiz . * 4
* , 114 ( 1966 ) ] .
e. waxman , astro - ph/9804023 .
s. yoshida , g. sigl and s. j. lee , phys .
* 81 * , 5505 ( 1998 ) .
j. j. blanco - pillado , r. a. vazquez and e. zas , phys .
d * 61 * , 123003 ( 2000 ) .
g. gelmini and a. kusenko , phys .
lett . * 82 * , 5202 ( 1999 ) .
g. gelmini and a. kusenko , phys .
* 84 * , 1378 ( 2000 ) .
t. j. weiler , hep - ph/9910316 .
j. l. crooks , j. o. dunn and p. h. frampton , astrophys .
j. * 546 * , l1 ( 2001 ) .
g. b. gelmini , hep - ph/0005263 .
h. ps and t. j. weiler , phys .
d * 63 * , 113015 ( 2001 ) .
d. fargion , m. grossi , p. g. de sanctis lucentini , c. di troia and r. v. konoplich , astro - ph/0102426 .
z. fodor , s. d. katz and a. ringwald , hep - ph/0105064 .
z. fodor , s. d. katz and a. ringwald , hep - ph/0105336 .
b. h. mckellar , m. garbutt , g. j. stephenson and t. goldman , hep - ph/0106123 .
a. ringwald , hep - ph/0111112 .
o. e. kalashev , v. a. kuzmin , d. v. semikoz and g. sigl , hep - ph/0112351 . g. gelmini and g. varieschi , hep - ph/0201273 .
g. l. fogli , e. lisi , d. montanino and a. palazzo , phys .
d * 64 * , 093007 ( 2001 ) .
j. n. bahcall , m. c. gonzalez - garcia and c. pena - garay , jhep * 0108 * , 014 ( 2001 ) .
m. c. gonzalez - garcia and y. nir , hep - ph/0202058 . c. weinheimer
_ et al . _ ,
b * 460 * , 219 ( 1999 ) . v. m. lobashev _ et al .
_ , phys .
b * 460 * , 227 ( 1999 ) .
j. bonn _ et al .
_ , nucl .
suppl . * 91 * , 273 ( 2001 ) .
v. m. lobashev _ et al .
_ , nucl .
suppl . * 91 * , 280 ( 2001 ) . c. e. aalseth _
[ igex collaboration ] , phys .
c * 59 * , 2108 ( 1999 ) .
l. baudis _ et al .
_ , phys .
* 83 * , 41 ( 1999 ) . c. e. aalseth _
et al . _ ,
phys . atom .
nucl . * 63 * , 1225 ( 2000 ) [ yad
. fiz . *
63 * , 1299 ( 2000 ) ] .
h. v. klapdor - kleingrothaus _ et al .
. phys . j. a * 12 * , 147 ( 2001 ) .
c. e. aalseth
_ et al . _ ,
hep - ex/0202026 . h. v. klapdor - kleingrothaus , a. dietz , h. l. harney and i. v. krivosheina , mod .
a * 16 * , 2409 ( 2002 ) .
f. feruglio , a. strumia and f. vissani , hep - ph/0201291 .
c. e. aalseth _
et al . _ , hep - ex/0202018 . v. d. barger , t. j. weiler and k. whisnant , phys . lett .
b * 442 * , 255 ( 1998 ) .
h. v. klapdor - kleingrothaus , h. pas and a. y. smirnov , phys .
d * 63 * , 073005 ( 2001 ) .
f. vissani , nucl .
suppl . * 100 * , 273 ( 2001 ) .
s. m. bilenky , s. pascoli and s. t. petcov , phys .
d * 64 * , 053010 ( 2001 ) .
s. m. bilenky , s. pascoli and s. t. petcov , phys .
d * 64 * , 113003 ( 2001 ) .
y. farzan , o. l. peres and a. y. smirnov , nucl .
b * 612 * , 59 ( 2001 ) .
p. osland and g. vigdel , phys .
b * 520 * , 143 ( 2001 ) .
m. czakon , j. gluza , j. studnik and m. zralek , phys .
d * 65 * , 053008 ( 2002 ) .
r. a. croft , w. hu and r. dave , phys .
lett . * 83 * , 1092 ( 1999 ) .
m. fukugita , g. c. liu and n. sugiyama , phys .
* 84 * , 1082 ( 2000 ) .
e. gawiser , astro - ph/0005475 .
x. wang , m. tegmark and m. zaldarriaga , astro - ph/0105091 .
t. j. loredo and d. q. lamb , in _
dallas 1988 , proceedings , relativistic astrophysics _ , pp . 601 - 630 .
p. j. kernan and l. m. krauss , nucl .
b * 437 * , 243 ( 1995 ) .
p. gondolo , g. gelmini , and s. sarkar , nucl .
b * 392 * , 111 ( 1993 ) .
d. e. groom _ et al .
_ [ particle data group collaboration ] , eur . phys .
j. c * 15 * , 1 ( 2000 ) .
d. buskulic _ et al . _
[ aleph collaboration ] , z. phys .
c * 66 * , 355 ( 1995 ) .
k. abe _ et al . _
[ sld collaboration ] , in _ proc . of the 19th intl .
on photon and lepton interactions at high energy lp99 _ , ed .
jaros and m.e .
peskin , hep - ex/9908033 .
opal pn299 , unpublished .
y. i. azimov , y. l. dokshitzer , v. a. khoze and s. i. troian , z. phys .
c * 27 * , 65 ( 1985 ) .
y. i. azimov , y. l. dokshitzer , v. a. khoze and s. i. troian , z. phys .
c * 31 * , 213 ( 1986 ) . c. p.
fong and b. r. webber , nucl .
b * 355 * , 54 ( 1991 ) .
t. sjostrand , comput .
commun . * 82 * , 74 ( 1994 ) .
j. n. bahcall and e. waxman , astrophys .
j. * 542 * , 543 ( 2000 ) .
z. fodor and s. d. katz , phys .
d * 63 * , 023002 ( 2001 ) .
e. waxman , astrophys .
j. * 452 * , l1 ( 1995 ) .
z. fodor , s. d. katz and g. papp , hep - lat/0202030 .
s. lee , phys .
d * 58 * , 043004 ( 1998 ) .
r. j. protheroe and p. a. johnson , astropart .
* 4 * , 253 ( 1996 ) .
r. j. protheroe and p. l. biermann , astropart .
* 6 * , 45 ( 1996 ) [ erratum - ibid . *
7 * , 181 ( 1996 ) ] .
t. a. clark , l. w. brown and j. k. alexander , nature * 228 * , 847 ( 1970 ) .
r. j. protheroe , nucl .
suppl . * 77 * , 465 ( 1999 ) .
r. gandhi , nucl .
suppl . * 91 * , 453 ( 2000 ) .
j. g. learned and k. mannheim , ann .
nucl . part .
sci . * 50 * , 679 ( 2000 ) .
r. m. baltrusaitis _ et al .
_ , phys .
d * 31 * , 2192 ( 1985 ) .
p. w. gorham , k. m. liewer , c. j. naudet , d. p. saltzberg and d. r. williams , astro - ph/0102435 .
e. waxman and j. n. bahcall , phys .
d * 59 * , 023002 ( 1999 ) .
j. n. bahcall and e. waxman , phys .
d * 64 * , 023002 ( 2001 ) .
k. mannheim , r. j. protheroe and j. p. rachen , phys .
d * 63 * , 023003 ( 2001 ) .
d. seckel _ et al . _
[ rice collaboration ] , in _ proc .
27th international cosmic ray conference _ , hamburg , germany , 2001 , pp .
1137 - 1140 .
v. s. beresinsky and g. t. zatsepin , phys .
b * 28 * , 423 ( 1969 ) .
berezinsky , v.s . and zatsepin , g.t . ,
j. nucl .
phys . * 11 * , 111 ( 1970 ) [ yad .
fiz . * 11 * , 200 ( 1970 ) ] .
s. yoshida and m. teshima , prog .
phys . * 89 * , 833 ( 1993 ) .
r. engel and t. stanev , phys .
d * 64 * , 093010 ( 2001 ) .
berezinsky , v.s . , in : _ proc .
dumand summer workshops _ , learned , j.g . ( ed . ) , khabarovsk and lake baikal , 22 - 31 aug 1979 , hawaii dumand center , university of hawaii , 1980 , pp .
245 - 261 .
v. berezinsky , m. kachelriess and a. vilenkin , phys .
lett . * 79 * , 4302 ( 1997 ) .
v. a. kuzmin and v. a. rubakov , phys . atom .
nucl . * 61 * , 1028 ( 1998 ) [ yad .
fiz . * 61 * , 1122 ( 1998 ) ] .
m. birkel and s. sarkar , astropart .
* 9 * , 297 ( 1998 ) .
s. sarkar and r. toldra , nucl .
b * 621 * , 495 ( 2002 ) .
v. berezinsky and m. kachelriess , phys .
b * 434 * , 61 ( 1998 ) .
v. berezinsky and m. kachelriess , phys .
d * 63 * , 034007 ( 2001 ) .
r. toldra , comput .
commun . * 143 * , 287 ( 2002 ) .
r. toldra , astro - ph/0201151 . c. barbot and m. drees , hep - ph/0202072 .
a. ibarra and r. toldra , hep - ph/0202111 .
v. s. berezinsky and a. vilenkin , phys .
d * 62 * , 083512 ( 2000 ) . c. p. ma , astro - ph/9904001 .
j. r. primack and m. a. gross , astro - ph/0007165 .
j. r. primack , private communication . l. n. da costa , w. freudling , g. wegner , r. giovanelli , m. p. haynes and j. j. salzer , astrophys . j. * 468 * , l5 ( 1996 ) .
a. dekel _ et al .
_ , astrophys . j. * 522 * , 1 ( 1999 ) .
j. p. kneller , r. j. scherrer , g. steigman and t. p. walker , phys .
d * 64 * , 123506 ( 2001 ) .
a. d. dolgov , s. h. hansen , s. pastor , s. t. petcov , g. g. raffelt and d. v. semikoz , hep - ph/0201287 . m. ave , j. a. hinton , r. a. vazquez , a. a. watson and e. zas , astro - ph/0110613 . k. shinozaki _ et al .
_ [ agasa collaboration ] , in _ proc .
27th international cosmic ray conference _ , hamburg , 2001 , pp .
346 - 349 .
s. l. dubovsky and p. g. tinyakov , jetp lett . * 68 * , 107 ( 1998 ) .
v. berezinsky and a. a. mikhailov , phys .
b * 449 * , 237 ( 1999 ) .
g. a. medina tanco and a. a. watson , astropart .
* 12 * , 25 ( 1999 ) . c. k. guerard , nucl .
* 75a * , 380 ( 1999 ) .
x. bertou , m. boratav and a. letessier - selvon , int .
j. mod .
a * 15 * , 2181 ( 2000 ) .
p. sreekumar _ et al .
_ , astrophys . j. * 494 * , 523 ( 1998 ) . s. barwick , talk at radhep 2000 , ucla workshop , www.ps.uci.edu/~amanda a. osipowicz _ et al .
_ [ katrin collaboration ] , hep - ex/0109033 . c. marquet [ nemo collaboration ] , nucl .
suppl . * 87 * , 298 ( 2000 ) .
a. alessandrello _ et al .
_ [ cuore collaboration ] , hep - ex/0201038 . h. v. klapdor - kleingrothaus , j. hellmig and m. hirsch , j. phys .
g * 24 * , 483 ( 1998 ) .
m. danilov _ et al .
_ , phys .
b * 480 * , 12 ( 2000 ) .
j. f. beacom , r. n. boyd and a. mezzacappa , phys .
lett . * 85 * , 3568 ( 2000 ) .
n. arnaud , m. barsuglia , m. a. bizouard , f. cavalier , m. davier , p. hello and t. pradier , phys .
d * 65 * , 033010 ( 2002 ) .
o. catalano , nuovo cim . * 24c * , 445 ( 2001 ) . | we consider the possibility that a large fraction of the ultrahigh energy cosmic rays are decay products of z bosons which were produced in the scattering of ultrahigh energy cosmic neutrinos on cosmological relic neutrinos .
we compare the observed ultrahigh energy cosmic ray spectrum with the one predicted in the above z - burst scenario and determine the required mass of the heaviest relic neutrino as well as the necessary ultrahigh energy cosmic neutrino flux via a maximum likelihood analysis .
we show that the value of the neutrino mass obtained in this way is fairly robust against variations in presently unknown quantities , like the amount of neutrino clustering , the universal radio background , and the extragalactic magnetic field , within their anticipated uncertainties .
much stronger systematics arises from different possible assumptions about the diffuse background of ordinary cosmic rays from unresolved astrophysical sources . in the most plausible case that these ordinary cosmic rays are protons of extragalactic origin ,
one is lead to a required neutrino mass in the range @xmath0 ev @xmath1 ev at the 68% confidence level .
this range narrows down considerably if a particular universal radio background is assumed , e.g. to @xmath0 ev @xmath2 ev for a large one .
the required flux of ultrahigh energy cosmic neutrinos near the resonant energy should be detected in the near future by amanda , rice , and the pierre auger observatory , otherwise the z - burst scenario will be ruled out . |
fiducial theory was introduced by @xcite to avoid the problems related to the choice of a prior distribution .
fiducial inference has not gained much popularity as such , but the related theory has been historically influential [ @xcite ] , and is still important in the current flow of statistical developments [ @xcite ] .
fisher s own view on fiducial inference evolved over the years as can be inferred from a reading of his initial [ fisher ( @xcite ) ] and more final formulation of the theory [ @xcite ] .
he was in particular very positive to the developments by fraser ( @xcite ) , and we most certainly share this point of view .
fraser ( @xcite ) develops the theory and uses the label _
structural inference _ for this .
a strongly related theory was presented under the label of _ functional models _ by @xcite and @xcite .
the term _ fiducial _ will here be used more generally so that it includes structural , functional , and the original fiducial arguments given by fisher .
the original idea of fisher was to obtain the fiducial distribution directly from the cumulative distribution , but this line of argument runs into problems when similar arguments are tried in the multivariate case .
the definition used here is based on the solution of a fiducial equation , and is in this sense similar to the approaches of @xcite , @xcite and hannig ( @xcite ) . a more precise definition of the term _ fiducial model _
as used here is given in section [ soptimal ] in definition [ deffidmod ] . a brief review of alternative , but strongly related definitions found in the literature is given in the final section [ sdiscussion ] .
let @xmath0 denote the realized loss for an action @xmath1 given the model parameter @xmath2 .
let @xmath3 be the sample space equipped with the @xmath4-field @xmath5 of events .
the risk @xmath6 of a decision rule @xmath7 is by definition equal to the expected value @xmath8 .
this is determined by the statistical model given by the family @xmath9 of probability spaces .
consider now the more special case where @xmath10 , possibly after a suitable change of variables .
assume that @xmath11 is a measurable quasigroup with a unit @xmath12 , and product @xmath13 written like ordinary multiplication [ @xcite ] .
this includes the more common case of a group , but it is more general since the associative law is not assumed to hold . assume furthermore that @xmath14 conditionally on @xmath15 and that the law of @xmath16 does not depend on @xmath17 .
this gives an example of a fiducial model for the statistical model as defined more generally on page .
the fiducial distribution is obtained by solving the fiducial equation @xmath18 for @xmath17 when @xmath19 is sampled from @xmath20 .
existence and uniqueness is ensured since @xmath11 is a quasi - group .
a variable @xmath21 is uniquely determined by @xmath22 .
the fiducial distribution is then the conditional law of @xmath21 given @xmath15 .
assume that the loss is invariant in the sense that @xmath23 , and that the decision rule is equivariant in the sense that @xmath24 .
the assumptions ensure the validity of the following calculation : [ eqriskcalc ] @xmath25 & = & { e}^\theta\gamma\bigl(\theta , \delta(\theta u)\bigr ) \hspace*{16.74pt}\qquad \mbox{fiducial model for } { \mathrm{p}}_x^\theta \\[-2pt ] & = & { e}^\theta\gamma\bigl(\theta , \theta\delta(u)\bigr ) \hspace*{16.95pt}\qquad\mbox { equivariance of $ \delta$ } \\[-2pt ] \label{eqrisk4 } & = & { e}^\theta\gamma\bigl(e , \delta(u)\bigr ) \hspace*{23pt}\qquad\mbox{invariance of $ \gamma$ } \\[-2pt ] & = & { e}^\theta\gamma\bigl(\theta^x , \theta^x \delta(u)\bigr ) \qquad \mbox{invariance of $ \gamma$ } \\[-2pt ] & = & { e}^\theta\gamma\bigl(\theta^x , \delta\bigl ( \theta^x u\bigr)\bigr ) \hspace*{1pt}\qquad \mbox{equivariance of $ \delta$ } \\[-2pt ] \label{eqqfid } & = & { e}^\theta\gamma\bigl(\theta^x , \delta(x)\bigr ) \hspace*{17.74pt}\qquad \mbox{fiducial equation}.\end{aligned}\ ] ] a variation of the above argument gives that @xmath21 can be replaced by @xmath26 in the conclusion . in the group case the law of @xmath21 will coincide with the law of @xmath26 , but in general not since the defining equation @xmath27 of the right inverse @xmath28 does not provide the solution of the fiducial equation .
it follows from this that an optimal equivariant rule , if it exists , is determined by the fiducial distribution of @xmath21 or by the distribution of @xmath26 from the right inverse .
the first part of the claims in the abstract has hence been established .
it is hoped that the reader can appreciate the simplicity and consequence of the calculation given in equation ( [ eqriskcalc ] ) , but it could also be considered to be essentially greek . the required theory of decisions and fiducial theory will be explained in some more detail in section [ soptimal ] , and examples are presented in section [ sexamples ] .
the presentation is essentially as given in standard textbooks [ @xcite ] , but with the simplifications given by a fiducial model .
the monographs by @xcite and @xcite are recommended as excellent sources for theory and examples beyond the standard textbooks .
the presentation in the following will be mostly restricted to the group case , but it will be more general than the previous in the sense that the assumption of equality of the involved spaces will be abandoned
. it will be more general than standard theory since , as above , the arguments will not depend on existence of invariant measures .
consider the case where the loss of an action @xmath29 is of the form @xmath30 corresponding to a statistical model @xmath31 .
it is here assumed that the model parameter @xmath32 is a @xmath4-finite random quantity and this and all other random quantities are defined based on the underlying conditional probability space @xmath33 as explained by @xcite .
this means in particular that @xmath34 , and @xmath35 , @xmath36 are measurable functions .
it means also that all expectations that occur are defined by integration over @xmath37 . as an example @xmath38 by definition .
it is here assumed that @xmath39 , @xmath40 , and @xmath41 are measurable .
the conditional law @xmath42 is well defined if @xmath43 is @xmath4-finite .
the consequence @xmath44 is a theorem .
the law @xmath45 of @xmath32 is not assumed known and is not needed in the arguments which follow .
the reason for the assumption of existence of @xmath32 , @xmath46 , and indeed any random quantity involved in the arguments , as functions defined on the conditional probability space @xmath47 is as in the formulation of probability theory given by @xcite : any collection of random quantities gives a new random quantity with a well - defined law , and measurable functions of random quantities give new random quantities .
the theory is completely based on the underlying abstract space @xmath37 . a _ group invariant problem _ is given by a group @xmath11 that has a transformation group action on the sample space @xmath3 , the model parameter space @xmath48 , and the action space @xmath49 .
the problem is group invariant if @xmath50 and @xmath51 .
an inference rule @xmath52 with a corresponding action @xmath53 is _ equivariant _ if @xmath24 .
the restriction to the class of equivariant actions can be interpreted as a consistency requirement : an observation @xmath54 from @xmath55 corresponds to an observation @xmath56 from @xmath57 .
the two corresponding problems are formally identical and the use of an equivariant action ensures consistency . the problem considered here is to determine an equivariant @xmath52 such that the risk @xmath58 is minimized .
it will be assumed that @xmath59 with the action given by the group multiplication @xmath60 directly .
the _ orbit _ of @xmath54 in @xmath3 is defined by @xmath61 , and likewise for orbits in @xmath48 and @xmath49 .
the action of @xmath11 is _ free _ on @xmath3 if the mapping @xmath62 is injective for all @xmath54 .
the group action is _ transitive _ on @xmath3 if @xmath63 .
if the group action is both transitive and free , then it is said to be _ regular _ and the corresponding space is then a principal homogeneous space for @xmath11 .
it follows in particular that the model parameter space @xmath48 is a principal homogeneous space for @xmath11 , but there has also been an identification of the identity element @xmath12 in @xmath48 .
let @xmath64 be a random quantity such that holds identically for all @xmath65 and @xmath17 .
it follows then that @xmath66 since the group invariance of the statistical model justifies @xmath67 .
this construction proves that @xmath68 with @xmath69 is a fiducial model for @xmath55 .
the concept of a fiducial model is defined as follows .
[ deffidmod ] let @xmath32 be a @xmath4-finite random quantity .
a fiducial model @xmath70 is given by a random quantity @xmath64 and a measurable function @xmath71 .
this is a fiducial model for the statistical model @xmath72 if @xmath73 the notation @xmath74 means that @xmath75 so definition [ deffidmod ] can be compared with equation ( [ eqgfid2 ] ) .
it is allowed in the above that @xmath20 does depend on @xmath17 .
interesting examples where this occurs are discussed by @xcite in the form of dependence on shape parameters in addition to pure group parameters . in the following it will , however , be assumed throughout that the fiducial model is _ conventional _ in the sense that @xmath20 does not depend on @xmath17 .
it is important to notice that many different fiducial models are possible for a given statistical model .
a fiducial model provides a different basis for statistical inference than a statistical model .
the choice of a particular fiducial model can be compared with the choice of a bayesian prior together with a statistical model .
fiducial inference is then initially different from frequentist and bayesian inference since the inferential basis is given by a fiducial model which is assumed known .
fiducial inference as such will not be considered here , but the corresponding fiducial algorithms will be used as vehicles for the construction of frequentist procedures .
a fiducial model @xmath70 is _ simple _ if the fiducial equation @xmath76 has a unique solution @xmath77 when solved for @xmath17 for all @xmath78 . in the simple and conventional case
the fiducial distribution is defined as the distribution of @xmath79 conditional on @xmath15 .
[ deffid1 ] let @xmath70 be a conventional simple fiducial model .
define the random quantity @xmath80 by @xmath81 .
the fiducial distribution is the conditional law of @xmath80 given @xmath15 .
the fiducial model @xmath68 given by equation ( [ eqgroupmod ] ) is simple if and only if @xmath3 is a principal homogeneous space for @xmath11 . in this case , by the choice of a unit element in @xmath3 , the identification @xmath82 can be done .
it follows then that @xmath83 is the unique solution of @xmath18 , and the fiducial distribution is the conditional distribution of @xmath84 as it appears in equation ( [ eqqfid ] )
. the remainder of this section will be on the analysis of the group model by means of the constructed fiducial model given by equation ( [ eqgroupmod ] ) in the case where @xmath3 is not assumed to be a principal homogeneous space for @xmath11 .
the aim is to determine an equivariant inference rule @xmath52 so that the risk given by equation ( [ eqrisk ] ) is minimized .
a definition of a fiducial distribution will also be presented for this group case .
the resulting distribution coincides with the distribution described with many more examples , explicit calculation of densities , and illustrative figures by fraser ( @xcite ) .
a first observation is that the calculations given by equations ( [ eqrisk1])([eqrisk4 ] ) are valid and the risk is given by @xmath85 .
the construction of the fiducial model has hence given a simple proof that gives that the risk does not depend on the model parameter since @xmath20 does not depend on @xmath17 .
let @xmath86 be an invariant statistic in the sense that @xmath87 for all @xmath88 .
this is equivalent with the requirement that @xmath89 is constant on all orbits in the sample space @xmath3 .
the fiducial model in equation ( [ eqgroupmod ] ) gives that @xmath90 conditionally on @xmath15 .
the conclusion is that @xmath91 does not depend on @xmath17 and has a known distribution .
this proves that an invariant statistic @xmath92 is ancillary .
assume furthermore that @xmath86 is a _
maximal invariant statistic_. this means that the family of level sets of @xmath89 coincides with the family of orbits in @xmath3 .
let @xmath54 be given and assume that @xmath93 .
the maximality ensures that @xmath94 , so @xmath95 for some @xmath96 .
this @xmath96 will be unique if @xmath11 acts freely on @xmath3 , but here it will more generally be assumed that @xmath96 is determined by the choice of a measurable selection .
the measurable selection theorem [ @xcite ] ensures existence under mild conditions .
the fiducial distribution of the corresponding variable @xmath21 can be described as follows .
[ deffid2 ] let @xmath19 be a sample from the distribution of @xmath97 where @xmath89 is a maximal invariant .
let @xmath96 be a measurable selection solution of @xmath95 .
this @xmath96 is a sample from a fiducial distribution .
the solution @xmath96 exists since @xmath93 ensures that @xmath54 and @xmath19 are on the same orbit .
definition [ deffid2 ] is a special case of definition [ deffid1 ] if @xmath3 is a principal homogeneous space for @xmath11 , and the definitions are hence consistent .
it is possible to define a fiducial distribution for more general cases .
one version is as presented by @xcite , but there are also other possibilities available .
this will not be discussed further here since the given definitions of the fiducial distribution are sufficient for the purposes in this paper .
let @xmath86 be a maximal invariant statistic .
the calculation that gave equation ( [ eqrisk4 ] ) can now be continued to give @xmath98 { \mathrm{p}}^\theta_y ( dy ) . \ ] ] the expression @xmath99 $ ] does only depend on @xmath100 .
the optimal rule @xmath52 , if it exists , is found by minimization for each given @xmath100 .
assume that @xmath54 is such that @xmath101 .
it follows then that @xmath102 and the optimal rule @xmath52 is determined by the fiducial distribution of @xmath21 .
the variable @xmath21 is defined as a measurable selection solution of @xmath22 .
this result can be summarized as the main technical result in this paper .
[ theo1 ] the risk of an equivariant rule in a group invariant problem is determined by a fiducial distribution if the model parameter space is a principal homogeneous space for the group .
it should be noted that the statement assumes existence of a fiducial distribution as described above , but uniqueness of a fiducial distribution is not assumed .
optimal inference procedures are determined by the fiducial distribution regardless of the choice of a measurable selection for the determination of a fiducial distribution .
the optimal @xmath52 is found , if it exists , as the minimizer @xmath103 of the expression @xmath104 where the conditional distribution of @xmath21 is a fiducial distribution .
theorem [ theo1 ] generalizes directly to the larger class of randomized equivariant actions .
this is obtained by a replacement of the equivariant action @xmath105 by the randomized equivariant action @xmath106 in the calculations .
it is here assumed that @xmath64 and @xmath107 are conditionally independent in the sense that @xmath108 , and both conditional distributions do not depend on @xmath17 .
the equivariance is defined by the identity @xmath109 .
a randomized action corresponds to the assignment of a probability measure on the action space @xmath49 .
the set of randomized actions is hence always a convex set , and this gives theoretical advantages to the problem of minimization of the risk .
if , however , the loss function @xmath110 is convex on @xmath49 for each @xmath17 , then the jensen inequality gives that it is sufficient to consider nonrandom actions [ @xcite , page 48 ] .
theorem [ theo1 ] generalizes also directly to the case where @xmath11 is only assumed to act transitively on @xmath48 .
the construction is as above , and starts with fixing a @xmath111 and the construction of a random variable @xmath64 such that @xmath112 .
all the arguments given above can then be repeated with @xmath11 playing the role of a new and possibly larger parameter space .
the result is then first a fiducial distribution on @xmath11 , but this is pushed forward to a fiducial distribution on @xmath48 by the mapping @xmath113 .
it is known that the fiducial coincides with the posterior from a right haar prior , and for these cases theorem [ theo1 ] is a known result with the posterior used in the formulation instead .
there are , however , groups where no haar prior exists , and in this case theorem [ theo1 ] and its extensions given by the above comments is a novel result .
the derivation given in the also gives a similar result in the more general case of a quasi - group , and the existence of invariant measures is then also not automatic .
the examples presented next are chosen to illustrate the concepts . many more examples and thorough discussions are found in the previously quoted textbooks and monographs . a complete treatment of the given examples including in particular simulation studies of the resulting procedures will not be pursued since this would tend to take attention away from the main issue .
the purpose is simply to indicate the usefulness of fiducial theory . a random sample of size n from the bernoulli distribution provides an example where the results related to theorem [ theo1 ] can not be applied directly .
fiducial theory can , nonetheless , be used to obtain optimal inference .
the largest possible group @xmath11 equals @xmath114 corresponding to the group of permutations of the set @xmath115 .
the action on @xmath116 is determined by @xmath117 , and the set of orbits in the parameter space is uncountable .
the conclusion is that conditioning on the maximal invariant as in the arguments leading to theorem [ theo1 ] does not provide any essential simplification of the problem .
this example is , however , very important from the point of view that fiducial distributions can still provide optimal procedures .
@xcite has constructed a randomized most powerful unbiased confidence interval , and this is related to a fiducial distribution [ @xcite , stevens ( @xcite ) ] .
the empirical mean is the unbiased estimator of @xmath118 with minimum variance .
it can , however , be argued that neither unbiasedness nor minimum variance are natural concepts in this particular case .
the parameter space @xmath48 can alternatively be identified with the circular arc @xmath119 in the @xmath120-plane .
this has the advantage that the fisher information metric distance between two distributions in this parametric family equals the distance along the arc [ @xcite ] .
the distance squared provides a loss that is invariant with respect to @xmath11 .
a natural task is to investigate on existence of an optimal equivariant estimator of @xmath118 with respect to the distance squared on the arc .
a reasonable candidate arises from the previously referenced fiducial distribution , but this will not be discussed further here .
the purpose here is to give an example which does not involve a group and where the argument given in equation ( [ eqriskcalc ] ) provides a fiducial distribution that can be used for the determination of the possibility of an optimal decision rule .
the octonions is here used as an example since it is one of the more interesting examples of a group - like structure where the associative law fails .
it has also a natural invariant loss that can be used in the arguments that follow .
a more familiar example without associativity can be constructed for the original model of fisher for the correlation coefficient , but we have not been able to identify a natural invariant loss in that case . the cayley
dickson construction defines a multiplication @xmath121 and an involution @xmath122 on @xmath123 where @xmath124 is an algebra with an involution . starting with the real numbers @xmath125 this gives the complex numbers @xmath126 .
repeated application of the construction gives then next the quaternions @xmath127 and then next the octonions @xmath128 . the octonions is hence equal to the @xmath129-dimensional vector space @xmath130 equipped with a particular multiplication operation so that @xmath128 is an algebra [ @xcite ] .
the number @xmath131 is the unit for multiplication , and every nonzero element @xmath54 has a multiplicative inverse @xmath132 with @xmath133 . the usual norm on @xmath130 is also given by the product and involution as @xmath134 , and the identity @xmath135 holds .
it follows in particular that @xmath136 .
the multiplication is not associative , but the algebra @xmath128 is alternative : the subalgebra generated by any two elements is associative .
consider next a fiducial model @xmath68 where @xmath137 is given by the product in @xmath128 , and where the conditional law @xmath20 is specified and does not depend on @xmath17 .
assume that @xmath138 where @xmath11 is a subset of @xmath128 that contains @xmath131 , the product of any two elements , and the inverse of any element .
the particular examples where @xmath11 is the nonzero octonions or where @xmath11 is the octonions with unit norm provide examples where @xmath11 is not a group since the associative law fails .
this is then a fiducial model for a statistical model @xmath139 where @xmath140 .
the corresponding fiducial distribution is the conditional distribution of @xmath84 given @xmath15 .
consider the case where the loss of an action @xmath141 is given by @xmath142 .
this loss is invariant , so the calculation in equation ( [ eqriskcalc ] ) gives that the risk of an equivariant decision rule is given by @xmath143 .
existence of an optimal estimator depends on @xmath20 or equivalently on @xmath21 , and this will not be discussed further here .
it can , however , be noted that any optimal equivariant decision rule is determined by @xmath144 , and @xmath145 is the minimizer of @xmath146 .
a rule on this form will be equivariant if @xmath145 belongs to the set @xmath147 @xmath148 .
there are many other examples of binary operations that are not associative .
a generic family of examples are produced by a relationship @xmath149 that has the property @xmath150 : it gives a one one correspondence between the domains of any two of the variables when the value of the third is fixed .
corresponding fiducial models based on @xmath151 defines the class of _ simple pivotal models _ in accordance with the terminology of @xcite , page 1057 .
concrete elementary examples are provided by @xmath152 , @xmath153 , and @xmath154 on suitable domains .
the property @xmath150 is conserved by a change of variables by one
one transformations resulting in @xmath155 .
for the given elementary examples , there exists a change of variables so that the result is a relation @xmath156 given by a group multiplication .
this is not possible in general .
simple counter examples arises for the fisherian simple pivotal models determined by the relation @xmath157 where @xmath158 is a suitable cumulative distribution function .
the prototypical example used by fisher [ ( @xcite ) , page 534 ] when he introduced the fiducial distribution is given by the sample correlation coefficient from a bivariate normal distribution .
in this case , a reduction to a group model as for the given elementary examples is not possible . in the general case starting from the property @xmath150 there exists , however , a change of variables that results in a relation given by a quasi - group with a unit : a loop .
the important conclusion of this short discussion is that the theory of simple pivotal models is linked naturally to the theory of loops .
the nonzero octonions provides an example of a loop which is not reducible to a group by a change of variables .
one purpose of this example is to demonstrate existence of a case where theorem [ theo1 ] can be used , but where an invariant measure does not exist .
let @xmath159 and @xmath160 where @xmath11 is a complex or real hilbert space .
the hilbert space @xmath11 is a group where the addition of vectors is the group operation , and an invariant loss is given by the squared distance between vectors as @xmath161 . a conventional fiducial model @xmath162 is given by @xmath163 for @xmath164 and a specification of a distribution @xmath20 that does not depend on @xmath17 . a maximal invariant is given by @xmath165 .
the fiducial distribution is given as the distribution of @xmath166 from the conditional law @xmath167 .
the optimal estimator of @xmath17 is given as @xmath168 .
it will be demonstrated in the next subsection that it is not necessary to assume independence of @xmath169 in the previous argument , and this assumption has indeed not been mentioned above .
more important is the fact that a right haar prior does not exist in the case where @xmath11 is an infinite - dimensional hilbert space .
an explicit example is given by @xmath170 .
the previous example has an infinite - dimensional parameter space , and this feature is quite common in applications as exemplified by nonparametric statistics . the example does also include data that are infinite dimensional , and this can be considered to be unrealistic in applications .
there are , however , applications where it is nonetheless convenient to assume that the observations are also infinite dimensional .
an important source of examples is given by the statistical signal processing literature [ @xcite ] .
explicitly , it can be convenient to assume that a signal is observed not only at discretely sampled points , but for all points .
similarly , it can be convenient to assume that a complete infinite sequence of sampled points is observed , even though only a finite number of samples are actually observed . in both cases
this can lead to a sample space that is not finite dimensional .
a related and very common convenience is to assume that observations are given by real numbers , even though the majority of concrete examples actually only involves a finite set of observable values due to limited instrument resolution [ @xcite ] .
explicit consideration of the limit from discretized data to continuous data gives , incidentally , a most promising route for the definition of fiducial distributions more generally than considered in section [ soptimal ] as demonstrated recently by @xcite .
if one observes the real random variables @xmath171 independently normally distributed with unknown mean @xmath172 and variance @xmath131 , it is customary to estimate @xmath173 by @xmath174 . if the loss is the sum of squares of the errors , this estimator is admissible for @xmath175 , but inadmissible for @xmath176 [ @xcite ] .
the optimal estimator derived above coincides with the customary estimator .
this exemplifies that the optimal estimator can be inadmissible .
the optimality is only ensured within the class of equivariant estimators .
equivariance can be a most natural demand , but this depends on the particular concrete modeling case at hand .
in certain situations [ @xcite ] it can be natural to give away the equivariance demand in order to obtain more precise estimates .
in other cases , especially in the context of physics , the equivariance demand can be closer to the foundation of the subject matter and will be an absolute demand . a particular case of the previous hilbert space example is given by assuming @xmath177 and where @xmath20 corresponds to a random sample of size @xmath178 from the uniform distribution on @xmath179 .
this gives then a fiducial model for a random sample from the uniform distribution on @xmath180 .
a fiducial distribution and a corresponding optimal estimator of @xmath17 follows from the hilbert space argument .
an alternative and more geometrically tractable argument follows as explained next from the use of the sufficient statistic given by the maximum and minimum observation .
let @xmath181 where the joint distribution of @xmath182 conditional on @xmath183 is given by the density @xmath184 on @xmath185 .
this is then a fiducial group model for the sufficient statistic given by the smallest and largest observation from a random sample from the uniform distribution on @xmath180 .
the model is also a special case of the hilbert space example with @xmath186 and where @xmath169 are conditionally dependent given @xmath187 .
reduction by sufficiency has here simplified the problem , but the fiducial equation is still over determined so a further reduction by the maximal invariant @xmath188 is necessary . the resulting conditional distribution @xmath189 becomes the uniform distribution on @xmath190 , and the fiducial distribution of @xmath21 becomes the uniform distribution on @xmath191 .
this is also a confidence distribution for @xmath17 .
the optimal estimator for @xmath17 given the invariant loss @xmath192 is @xmath193 .
we choose to add a few comments on this model and estimator since it has some unusual features .
a first observation is that the fisher information metric fails to exist due to nonexistence of the required derivative .
the corresponding distance between two distributions can , however , still be defined through the length of the parametric curve @xmath194 in the hilbert space of square integrable functions .
this curve is continuous , but the length from @xmath195 to @xmath196 is infinite : it is larger than @xmath197 for any integer @xmath178 .
the squared distance @xmath198 is the squared distance from the fisher information metric for any location family where the density is smooth .
based on this we consider the invariant loss @xmath192 to be a natural choice also in the nonsmooth example considered here .
the optimal estimator @xmath52 found above is unbiased and has hence minimum variance in the class of unbiased and equivariant estimators .
nonetheless , according to lehmann and casella [ ( @xcite ) , page 87 ] , there exists no uniformly minimum variance unbiased estimator of @xmath17 .
the statistic @xmath199 is a minimal sufficient statistic , but it is not complete .
the estimator @xmath52 is , however , the uniformly minimum variance unbiased estimator in the larger parametric family which also includes a scale parameter [ @xcite , vol .
2 , page 292 ] .
this later reference is also a very good source for further references and peculiarities regarding the uniform law.=-1 the following example is a scale example , and can be reduced to be a special case of the hilbert space location problem by the logarithmic transformation .
a direct solution is equally elementary and is presented to illustrate the derivation of an optimal estimator .
the explicit formula for the estimator is possibly a novelty .
assume that @xmath200 is a random sample of size @xmath178 from the exponential distribution with scale parameter @xmath201 .
a fiducial model is given by @xmath202 where the law @xmath203 is as for a random sample from the standard exponential distribution .
the arithmetic mean @xmath204 is a minimal sufficient statistic . a corresponding fiducial model is given by @xmath205 where @xmath206 is the law of a gamma variable with scale equal to @xmath207 and shape equal to @xmath178 .
this follows from well - known properties of the gamma distribution .
the model is both simple and conventional , and the fiducial distribution for an observation @xmath208 is hence the conditional distribution of @xmath209 .
the conclusion is that the fiducial distribution is the inverse - gamma with scale @xmath210 and shape @xmath178 . a direct but more lengthy calculation of the bayesian posterior corresponding to the right haar prior @xmath211 gives a posterior that coincides with the fiducial distribution found here .
it is well known more generally that the bayesian posterior from a right haar prior in a group model coincides with the fiducial .
the calculation demonstrates then that a fiducial model and the solution of the fiducial equation gives an alternative and in many cases simpler route for the calculation of the bayesian posterior .
the multivariate normal gives another example where the fiducial calculation is done in a few lines , but the corresponding bayesian calculation is much more cumbersome .
an added advantage of the fiducial calculation is that it shows directly that the corresponding fiducial distribution is a confidence distribution .
this is not easily obtained from the bayesian calculation .
the confidence distribution can alternatively be found by the likelihood ratio test , and this has the advantage of giving proof of optimality and corresponding optimal choices of confidence interval endpoints .
an alternative approach is to also derive optimal intervals based on theorem [ theo1 ] as exemplified by @xcite .
an alternative fiducial calculation can be done without the reduction to the complete sufficient statistic .
a maximal invariant @xmath89 is given by @xmath212 .
the conditional law @xmath213 will be concentrated on the ray @xmath214 with a distribution from a density for @xmath215 proportional to @xmath216 .
the assumption of a random sample from the standard exponential gives a particularly simple @xmath217 , and the fiducial is found explicitly as before .
the alternative calculation has the advantage that it can be used in the more general case where reduction by sufficiency is not available .
consider now the problem of estimation of @xmath218 with a loss given by @xmath219 .
this loss is a natural generalization of the squared error loss , but with the ordinary distance replaced by the distance @xmath220 which is the distance given by the fisher information metric in the case of the given scale model . in this case , @xmath221 with multiplication as the group operation .
the loss is equivariant , and it follows that the optimal rule @xmath52 based on the sufficient statistic @xmath46 is given as the minimizer of @xmath222 .
this gives that the optimal rule is determined from @xmath223 .
evaluation of the corresponding integral gives an explicit formula for the optimal rule .
it is @xmath224 where @xmath225 is the digamma function .
the estimator given by equation ( [ eqscalelossopt ] ) is possibly known in some contexts , but we have not found this explicit expression in any of the textbooks in the list of references or elsewhere .
assume that @xmath200 is a random sample of size @xmath178 from the gamma distribution with scale parameter @xmath201 and shape parameter @xmath215 .
the model parameter is @xmath226 .
this gives an example as in section [ ssbernoulli ] where the results related to theorem [ theo1 ] can not be applied directly .
fiducial theory can be used to obtain candidates for good frequentist inference procedures as indicated next .
particular results include an exact joint confidence distribution for @xmath227 , an exact confidence distribution for @xmath215 , and a recipe which produces estimators for functions of @xmath227 that depends on the choice of a loss .
a fiducial model is given by @xmath228 , where the law @xmath20 corresponds to a random sample of size @xmath178 from the uniform distribution on the unit interval @xmath179 , and @xmath229 is the inverse cumulative distribution function of a gamma variable with shape @xmath215 and scale @xmath131 .
let @xmath230 where @xmath231 and @xmath232 are the arithmetic and geometric means .
the bartlett statistic @xmath233 depends only on @xmath215 , and is independent of @xmath231 as a consequence of the basu theorem .
a corresponding fiducial model @xmath162 for @xmath55 is given by @xmath234 and @xmath235 .
it can be noted that @xmath236 gives separately a fiducial model for @xmath237 .
the corresponding fiducial distribution for @xmath215 is hence a confidence distribution .
an alternative fiducial model @xmath238 for @xmath237 is given by inversion of the cumulative distribution function for @xmath239 .
an alternative to @xmath240 is given by inversion for a gamma variable with shape @xmath241 and scale @xmath207 .
the combination gives an alternative fiducial model @xmath242 for @xmath55 with the property that @xmath243 defines a one one correspondence between any two variables when the third is kept fixed .
the law @xmath244 is the uniform law on the unit square @xmath245 .
coordinate transformations can be used to identify @xmath246 as sets with a quasigroup structure with a unit . both fiducial models @xmath162 and @xmath242 are simple and conventional and determine a fiducial distribution . for concreteness let @xmath21 be the fiducial corresponding to @xmath242 .
the quasigroup structure ensures that @xmath21 gives a joint exact confidence distribution for @xmath247 .
consider the problem of estimation of a function @xmath248 of the model parameter @xmath249 .
it can be allowed that @xmath250 is vector valued , but assume that each component is positive .
three examples that are included are then given by @xmath251 , @xmath252 , and @xmath253 .
a possible loss in these three cases is given by squared error loss on a logarithmic scale .
a candidate estimator @xmath52 is then given naturally by @xmath254 this can be evaluated by monte carlo simulation from @xmath244 which is the uniform distribution on the unit square @xmath255 ^ 2 $ ] .
another possibility is given by squared distance loss defined by the fisher information metric on @xmath48 in the case @xmath256 .
the foundations of bayesian and frequentist modeling and inference are well established both in terms of mathematical theory and interpretation .
we do not think that the same can be said about fiducial theory , but some readers may object to this .
a brief discussion of alternative formulations and naming conventions found in the literature seems hence to be in place .
definition [ deffidmod ] identifies a fiducial model with a pair @xmath70 .
the fiducial model is by definition _ conventional _ if @xmath20 does not depend on @xmath17 . in this case
we suggest to denote @xmath64 as the _ monte carlo variable _ and the measurable function @xmath257 as the _ fiducial relation_. the corresponding equation @xmath258 is the fiducial equation , but it may also equivalently be denoted as the fiducial relation .
a fiducial model @xmath70 is hence defined by a monte carlo variable @xmath64 and a fiducial relation @xmath257 . if @xmath19 is a sample from the monte carlo distribution @xmath20 , then @xmath259 is a sample from the statistical model as in definition [ deffidmod ] .
the inversion method gives the prototypical example with @xmath260 and @xmath20 equal to the uniform law on the interval @xmath255 $ ] .
this gives the link to the original definition of fisher , and also a justification of the choice of the term _ monte carlo variable _ since this represents a standard method for simulation from a statistical model on a computer .
the ingredients above given by the pair @xmath70 are also the starting point for dempster
shafer theory [ @xcite ] .
@xcite refer to @xmath64 as the _ auxiliary variable _ and the probability measure @xmath261 as the _ pivotal measure _ , where @xmath262 .
the equation @xmath263 is denoted the _ a - equation_. the whole set - up is referred to as an _ inferential model _ , and this is identified as something which is not equivalent to a statistical model . except for differences in naming conventions
it can be concluded that an _
inferential model _ is essentially the same as a _ conventional fiducial model _ as summarized in the previous paragraphs .
the dempster shafer calculus gives an alternative route for inference based on a fiducial model , but this is not discussed further here .
the discussion of fiducial theory we have presented is close to the presentation given by @xcite .
dawid and stone [ ( @xcite ) , page 1055 ] use the term _ fiducial model _ for the combination of @xmath79 and @xmath264 , and use the term _ functional model _ to describe the more general relation @xmath263 .
we chose to avoid the term _ functional model _ since the term _ functional data analysis _ is now the name of a branch of statistics .
@xcite denote the variable @xmath64 as the _ error variable _ , and uses the symbol @xmath265 instead .
this corresponds to the naming convention used by @xcite .
fraser [ ( @xcite ) , page 50 ] uses the terms _ structural model _ and _ structural equation _ in the case where group theory is an essential ingredient .
@xcite uses the term _ structural equation _ in stead of the term _ fiducial equation _ as used by us .
we have avoided the term _ structural _ here since there is an active and well - developed different theory which goes under the label of _ structural equations modeling_. our preference for the term _ fiducial _ as used here is mainly based on economy of language , and since this gives the direct link to the original papers of fisher .
the mathematically inclined reader may claim that definition [ deffidmod ] is not precise .
this , and the fact that this definition is a novelty compared with previous writers , motivate us to state in more detail the assumptions that are taken as implicitly given from the context in the statement of definition [ deffidmod ] .
the main difference is that every concept is based on an underlying abstract conditional probability space @xmath33 as stated initially in section [ soptimal ] .
the fiducial relation @xmath257 is a measurable function @xmath71 .
this means , as usual , that @xmath266 is a measurable set in the product @xmath4-algebra of @xmath267 whenever @xmath65 is a measurable set in @xmath268 .
a consequence is that @xmath269 is a random element in @xmath268 defined by the mapping @xmath270 .
this is measurable since it is assumed that @xmath36 and @xmath271 are measurable .
the conditional law @xmath20 of the monte carlo variable @xmath64 is known and does not depend on @xmath17 in the case of a conventional fiducial model .
if the considerations were limited to the case where @xmath33 is a probability space , then this would imply @xmath272 .
this fails generally as explained in more detail by @xcite since @xmath20 is a probability measure , but @xmath273 is unbounded if @xmath274 is unbounded .
the reason for allowing unbounded measures is the need to include improper priors @xmath45 .
this has proved fruitful in related ongoing research by the authors .
it gives in particular natural conditions that imply equality of bayesian posteriors and fiducial distributions . in specific modeling cases
the spaces @xmath275 , @xmath268 , @xmath48 , the fiducial relation @xmath257 , and the conditional law @xmath20 are all explicitly given .
this is as demonstrated by the examples in section [ sexamples ] .
the other ingredients mentioned above are not given explicitly since they rely on the underlying unspecified space @xmath37 .
this is as in the ordinary formulation of probability theory by @xcite where the whole theory is built upon the underlying abstract space @xmath37 .
existence must be proved in each specific modeling case , but follows trivially in many cases from the construction of a suitable product space .
optimal inference for the scale , the location , and the location - scale problems were investigated using fiducial theory by @xcite .
his presentation is most readable and is a good alternative to the presentations found in standard textbooks .
it can , however , be noted that he concludes that the confidence and the fiducial theories are essentially the same .
this is in contrast to the views of neyman and fisher .
they seemed to agree that in principle the fiducial distribution as described by fisher is not connected to the concept of confidence intervals as described by neyman and co - workers .
the content and aims of these two theories are different .
it seems clear that fisher never intended to get confidence intervals as the result of his fiducial arguments .
it is true that the fiducial distributions found in the location - scale problems , and more general group problem as in theorem [ theo1 ] , are confidence distributions , but we do consider the concepts to be essentially different in general .
the interpretation of the fiducial distribution , according to fisher [ ( @xcite ) , pages 54 and 59 ] is identical with the interpretation of the bayesian posterior : it represents the state of knowledge regarding the model parameter as a result of the model assumptions and the observation in the experiment .
it follows then in particular that the fiducial distribution of a function @xmath276 of the model parameter @xmath17 equals the distribution of @xmath277 where @xmath21 has the fiducial distribution .
this property does not hold for confidence distributions in general .
in addition , the fiducial distribution for a simple fiducial model as in definition [ deffid1 ] is not a confidence distribution in general [ @xcite ] .
the possibly most famous fiducial distribution is the fiducial distribution of the difference of means @xmath278 corresponding to two independent samples from two different normal distributions .
this fiducial distribution gives fisher s solution to the behrens fisher problem , but it can be shown by simulation that it is not a confidence distribution in the sense of having exact coverage probabilities .
a more general class of confidence distributions is defined by requiring not exact but conservative coverage probabilities .
this is in conformity with the definition of confidence sets in general .
exactness is often misguidedly taken as a measure of goodness , but it is not .
power of the associated tests gives one natural measure of goodness .
examples demonstrate that this more general concept of a confidence distribution does not coincide with fiducial distributions in general , but it seems to be an open question whether the behrens fisher fiducial distribution is a confidence distribution in this more extended sense .
numerical simulations indicate that the claim holds [ @xcite , page 269 ] .
the more general problem of obtaining a confidence interval for the linear combination of several means from different normal distributions is of considerable practical importance [ @xcite ] .
the iso recommended solution is in terms of a welch
satterthwaite solution , but a continuation of the arguments given by @xcite leads to the conclusion that the fiducial solution is a most competitive alternative solution .
the main virtue of the location - scale models in the context here is that they illustrate very well the reduction given by a maximal invariant in cases where a reduction by sufficiency is not possible .
this is also true for the multivariate models treated by @xcite . in this case
the multivariate normal can be reduced by sufficiency , but more general models can again be treated by a reduction through maximal invariants .
it seems that optimal , or good , inference procedures in these multivariate cases deserves further study guided by fiducial theory .
a recent example of this is given by @xcite , but there are a multitude of different possible examples as indicated by @xcite .
the suggestion given by theorem [ theo1 ] is that not only confidence intervals , but also other kinds of inference such as estimation should be considered .
eaton [ ( @xcite ) , pages 8991 ] considers the estimation of the covariance matrix from a multivariate normal sample .
he gives two possible candidates to use as a loss @xmath279 .
this exemplifies that in the multivariate cases , and in more complicated group cases , it can be difficult to decide upon which equivariant loss to use .
it can even be difficult to come up with a good candidate . in our examples
, it has been indicated that the squared distance from the fisher information metric is a natural choice
. this will be invariant under mild conditions .
for a statistical model @xmath280 , the distance is defined via the length of paths @xmath281 in the hilbert space @xmath282 . the nonparametric case given by a parameter space equal to all densities with respect to @xmath261
gives the distance @xmath283 .
the other end of the scale is given by a smooth finite - dimensional parametric model .
in this case , the previous leads to the fisher information metric : @xmath284 where @xmath285 . in either case
, it gives the model parameter space as a manifold equipped with a distance derived intrinsically from the statistical model .
the focus of fiducial theory has initially and currently most often been on the fiducial distribution by itself and the related possibility of construction of approximate or exact confidence intervals . the relation to other kinds of optimal inference such as estimation or prediction
was considered by hora and buehler ( @xcite ) .
the proofs they presented rely on the existence of an invariant measure , and it was clear that the fiducial in the case they considered corresponded to a bayesian posterior from the right haar prior . since then it has been established in a variety of problems that the bayesian algorithm can be used quite generally to obtain good or optimal frequentist procedures .
the calculation given in equation ( [ eqriskcalc ] ) can be taken as a strong indication that the fiducial algorithm can be used similarly to not only obtain confidence intervals , but also possibly good or optimal frequentist procedures more generally .
this statement is too general to be provable , but we consider nonetheless this to be the main content and message in this paper .
the point of view in this paper does not rely on any particular interpretation of the fiducial .
it is here simply viewed as a very convenient vehicle for the derivation of good , and sometimes optimal as in theorem [ theo1 ] , frequentist inference procedures .
we acknowledge the comments from the reviewers and associate editor , which resulted in an improved paper . | it is shown that the fiducial distribution in a group model , or more generally a quasigroup model , determines the optimal equivariant frequentist inference procedures .
the proof does not rely on existence of invariant measures , and generalizes results corresponding to the choice of the right haar measure as a bayesian prior .
classical and more recent examples show that fiducial arguments can be used to give good candidates for exact or approximate confidence distributions .
it is here suggested that the fiducial algorithm can be considered as an alternative to the bayesian algorithm for the construction of good frequentist inference procedures more generally . |
a longitudinal wave travelling through a fluid imparts a net drift motion to the suspended particles - an effect known as _
drift_. the _ classical _ stokes drift @xcite refers only to the deterministic behavior , i.e. it does not account for the stochastic fluctuations or perturbations in the system .
it has been extensively studied in various practical contexts like , for instance , the motion of tracers in meteorology and oceanography @xcite and that of the ( doping ) impurities in crystal growth @xcite .
a simple intuitive explanation of classical stokes drift is that the suspended particles spend a longer time in the regions of the wave - train where the force due to the wave acts in the direction of wave s propagation than the time spent in those regions where the force acts in the opposite direction ; therefore particles are driven in the sense of wave s propagation .
when several linearly superposed waves are present , the resulting drift velocity is simply the sum of the contributions from each wave @xcite . recently , there has been some interest in _ stochastic _
stokes drift ( @xcite , see also @xcite ) .
it was found that the diffusive motion of the suspended particles can significantly alter the amplitude of the drift velocity .
but it is the merit of @xcite to open stochastic stokes drift to further investigation by putting it in its natural context , namely _ ratchet - like brownian motors _ in which the thermal motion of small ( microscopic ) particles is rectified by an asymmetric time - dependent potential @xcite .
this scenario is referred to as the _ flashing ratchet_. paper @xcite is limited to the case when the motion of the overdamped suspended particles is diffusion - dominated , i.e. when the deterministic forcing due to the wave is very weak .
an approximate perturbative technique is developed for calculating the drift velocity in arbitrary dimensions , which shows that the drift depends on particles diffusivity in magnitude and , in more than 1d , also in direction ( a very appealing feature , useful in particle selection devices - see , for example , in @xcite doering et all , 1994 , and in @xcite m. bier , 1996 ) .
papers @xcite and , in a very different context @xcite , give an exact integral expression for stokes drift of overdamped diffusive particles in 1d , for arbitrary wave forms . using a fokker - planck equation technique
, it finds that white noise reduces stokes drift by comparison to its classical ( deterministic ) value , the flux reversal being absent .
the purpose of this short communication is to show that the characteristics of the 1d stokes drift may become more complex ( including flux reversal or enhancement of the classical drift ) when the particles are subjected to stochastic forcing that is non - white ( i.e. it has a finite correlation time ) .
we will focus on the case of dichotomic markov forcing , in which the particle is subjected to a randomly alternating force , in addition to the force exerted by the wave . for this choice of coloured noise , explicit analytic results
can be obtained , while this is not the case for other types of coloured noise , including the ornstein - uhlenbeck process . even though the time - average of the dichotomic force is taken to be zero ,
its effect in conjunction with the asymmetry of the wave s perturbation is highly nontrivial .
as such the model realizes in the stokes drift context another paradigm of brownian motors , namely that of the so - called _
rocking ratchet_. in this scenario , the systematic motion acquired under influence of an alternating zero - average force essentially revolves on the asymmetry of the nonlinear response in the presence of a steady but asymmetric potential @xcite .
the dichotomic markov forcing presents another important advantage in that the exact integral expression for stokes drift can be derived .
furthermore , in some cases of interest , the resulting integrals can be evaluated explicitly .
our analytic results are also complemented with computer simulations , which are found to be in full agreement with the theory .
the starting point is the following stochastic differential equation : @xmath0 where @xmath1 is a symmetric dichotomic noise ( a two - step markov process ) @xcite .
it can take only two values , @xmath2 , with equal probability , jumping between them with a probability @xmath3 per unit time .
it has zero mean and its autocorrelation function @xmath4 shows a finite correlation time @xmath5 . with a suitable time scaling , eq.([eq1 ] ) models the 1d overdamped motion of a brownian particle subjected to a dichotomic forcing .
@xmath6 is the particle position at time @xmath7 , while @xmath8 is the periodic forcing due to the wave travelling at the speed @xmath9 with wavelength @xmath10 : @xmath11 with @xmath12 as the case of physical relevance .
the quantity of interest is the _ drift velocity _ , which represents the mean velocity of the particles in the long time limit : @xmath13 where the mean is to be taken over the realisations of the dichotomic noise . by defining a new variable @xmath14
one can rewrite eq.([eq1 ] ) as : @xmath15 with @xmath16 , and therefore : @xmath17\ ; . \label{eq9}\end{aligned}\ ] ] here @xmath18 represent the probability densities for the particle position @xmath19 and the value of the dichotomic noise equal to @xmath2 , respectively .
their time evolution is described by @xcite : @xmath20 -k(\rho_{+}-\rho_{- } ) , \nonumber\\ \frac{\partial \rho_{-}(y , t)}{\partial t } & = & -\frac { \partial}{\partial y}[(f - a)\rho_{- } ] -k(\rho_{-}-\rho_{+ } ) .
\label{eq10}\end{aligned}\ ] ] the first term in the r.h.s . of each of these equations corresponds to the deterministic flow in phase space ( between two jumps of the dichotomic noise ) , while the other term corresponds to the jumps of the dichotomic noise between @xmath2 . because of the periodicity ( [ eq3 ] ) of @xmath21 ( @xmath22 ) , in order to evaluate ( [ eq9 ] ) we only need to study the asymptotic behavior of the reduced probability density : @xmath23 .
\label{eq11}\ ] ] from its definition , it is obvious that the reduced probability density is subjected to periodic boundary conditions : @xmath24 while the normalization condition reads :
@xmath25 considering also the quantity @xmath26\ ; , \label{eq14}\ ] ] it follows from ( [ eq10 ] ) that @xmath27 and @xmath28 obey a set of coupled evolution equations : @xmath29 the stationary solution @xmath30 can be easily found @xcite : @xmath31\ , dy ' , \label{eq23}\ ] ] which is valid as long as @xmath32 , @xmath33 . here
@xmath34 is defined by : @xmath35}{f^2(y')-a^2}dy ' . \label{eq20}\ ] ] and @xmath36 is the normalization factor according to eq.([eq13 ] ) .
therefore the drift velocity can be finally expressed as : @xmath37 these results are valid for an arbitrary wave form , provided that the obvious necessary differentiability and integrability conditions are fulfilled .
note that the drift depends nonlinearly on @xmath38 , so that the contributions of different waves are not additive .
an explicit analytical expression for @xmath30 can be obtained in the case of a piecewise linear wave forcing : @xmath39 , & y \in [ \lambda/2 - 2\varepsilon,\lambda/2 [ \ ; \mbox{mod } \;\lambda \\ - ( 1+b)v , & y \in [ \lambda/2,\lambda-2\varepsilon [ \ ; \mbox{mod } \ ; \lambda \\ - [ ( 1+b)v -bv(y-\lambda+2\varepsilon)/\varepsilon ] , & y \in [ \lambda-2\varepsilon,\lambda [ \ ; \mbox{mod } \ ; \lambda , \end{array } \right
. \label{eq26}\ ] ] with @xmath40 .
a square wave corresponds to @xmath41 , while @xmath42 represents the triangular - like wave .
the final expressions for @xmath30 are rather lengthy and will not be reproduced here .
but we performed numerical simulations , which show very good agreement with theoretical results , cf .
( [ fig : py ] ) . note that these results are valid for @xmath43 or for @xmath44 .
the analytic results for the square - like wave , corresponding to the limit @xmath45 , are of more interest .
they read as follows : @xmath46 \frac{2a^2b}{(1-b)^2v^2-a^2}\right\}\ ; , \label{eq28}\ ] ] for @xmath47 , and @xmath48 \frac{2a^2b}{(1+b)^2v^2-a^2}\right\}\ ; , \label{eq29}\ ] ] for @xmath49 , where we introduced the dimensionless variables : @xmath50}\ ; , \nonumber\\ v & \equiv & \frac{{\alpha}(1+b)}{2[(1+b)^2{\beta}^2 - 1]}\ ; , \nonumber \\ { \alpha } & \equiv & \frac{2kv\lambda}{a^2},\ \ \ \ \ { \beta}\equiv \frac{v}{a } \label{eq30}\end{aligned}\ ] ] stationary probability density for the piecewise linear wave , eq .
( [ eq26 ] ) , with @xmath51 , @xmath52 , @xmath53 and @xmath54 ( the arrows indicate the boundary points of the piecewise linear regions ) .
dashed lines represent the theoretical results , while solid lines correspond to simulations of @xmath55 particles evolving for 200 time units .
the insets show the two realisations of @xmath56 .
[ fig : py](a ) : @xmath57 , @xmath58 .
[ fig : py](b ) : @xmath59 , @xmath60.,scaledwidth=90.0% ] the following expression for the drift velocity is obtained : @xmath61 where @xmath62 is the classical value of the stokes drift velocity .
we mention some limiting cases of interest .
first , the limit @xmath63 or @xmath64 ( i.e. @xmath65 and/or @xmath66 ) leads to the classical value of the stokes velocity @xmath67 .
second , the drift velocity derived in @xcite is recovered in the white noise limit @xmath68 , @xmath69 , with @xmath70 finite ( @xmath71 ) . finally , the quenched - noise limit @xmath72 ( @xmath73 ) ( i.e. half of the particles , chosen at random , are subjected to a constant external forcing @xmath74 , while the other half to an external forcing @xmath75 ) results in the following mean velocity : @xmath76 this expression reveals the possibility of a _ flux reversal _ : the drift velocity is negative for @xmath77 , i.e. for sufficiently large amplitude . turning to the general case , we have represented in fig .
[ fig : drift ] the results of eq .
( [ eq31 ] ) together with simulations for the square wave for two representative sets of parameter values . for small values of
the forcing amplitude @xmath78 one finds that the drift is enhanced ( @xmath79 ) , but the enhancement is reduced upon increasing the frequency of the perturbation .
in fact one can prove that enhancement occurs whenever @xmath80 , by noting that @xmath81 and making use of the inequality @xmath82 .
on the other hand , a reduction of the drift or even a flux reversal ( @xmath83 ) can be induced at larger amplitudes ( for small @xmath3 , according to eq .
( [ eq36 ] ) ) , while the flux is now found to be an increasing function of the perturbation frequency .
in conclusion , the rocking ratchet version for stokes drift reveals two new features , which are to be attributed to the colour ( finite correlation time ) of the applied force : the drift is enhanced by small amplitude perturbations , while a flux reversal occurs at sufficiently large amplitude and large correlation time . while our calculations are limited to the case of dichotomic markov noise , we expect that similar results hold for other coloured noises . in
@xcite it was reported that , in two dimensions , both the direction and amplitude of the stokes drift can change with the intensity of the brownian motion when several waves are present .
a similar phenomenon will appear here when one considers , for example , two waves propagating in orthogonal directions @xmath84 and @xmath85 .
a dichotomic forcing @xmath86 will induce different drifts along these directions depending on its projections @xmath87 and @xmath88 , as well as on its transition rate @xmath3 .
in particular , by varying @xmath3 one is able to induce a significant change in the resulting drift direction , as can be seen in fig .
[ fig:2d ] for the simple example of identical square waves .
drift velocity according to eq .
( [ eq5 ] ) for a square wave with the same values of the parameters as in fig .
[ fig : py ] , except here @xmath89 .
symbols represent simulations with @xmath90 particles measured after @xmath91 time units .
the lines represent the theoretical value , eq .
( [ eq31 ] ) .
[ fig : drift](a ) : drift enhancement .
[ fig : drift](b ) : flux reversal.,scaledwidth=90.0% ] the resulting drift velocity ( in units @xmath92 ; thick arrows ) for two identical square waves ( with the same parameters as in fig .
[ fig : py ] ) propagating respectively along the @xmath93 and @xmath94 axes , in the presence of a stochastic forcing with @xmath95 , @xmath96 ( applied along the direction of the thin arrows ) , for different transition rates @xmath3 .
the dashed arrow corresponds to the classical limit @xmath97.,scaledwidth=65.0% ] g. g. stokes , trans .
* 8 * , 441 ( 1847 ) ; o. m. phillips , _ the dynamics of the upper ocean _ , cambridge university press , cambridge , 1977 ; j. lightwill , _ waves in fluids _ , cambridge university press , cambridge , 1978 . a few recent papers from a rich bibliography on the subject : + a. m. bratseth , tellus * 50a * , 451 ( 1998 ) ; s. m. cox , fluid dyn . res . *
19 * , 149 ( 1997 ) ; h. c. graber , b. k. haus , r. d. chapman and l. k. shay , j. geophys .
102 * , 18749 ( 1997 ) ( experimental ) ; j. c. mcwilliams , p. p.
sullivan and chin - hoh moeng , j. fluid mech .
* 334 * , 1 ( 1997 ) ; m. nordsveen and a. f. bertelsen , int . j. multiphase flow * 23 * , 503 ( 1997 ) ; v. polonichko , j. geophys . res . *
102 * , 15773 ( 1997 ) ; s. a. thorpe , j. phys . oceanogr .
* 27 * , 62 and 2072 ( 1997 ) .
m. o. magnasco , phys .
lett . * 71 * , 1477 ( 1993 ) ; m. bier , phys .
a * 211 * , 12 ( 1996 ) ; p. hnggi and r. bartussek , in _ lect .
notes in phys .
_ , vol . * 476 * , ed .
by j. parisi et al . ,
springer , berlin , 1996 ; r. d. astumian , science * 276 * , 917 ( 1997 ) ; f. julicher , a. ajdari and j. prost , rev .
phys . * 69 * , 1269 ( 1997 ) . | we derive the explicit analytic expression for the stokes drift in one dimension in the presence of a dichotomic markov forcing . for small amplitudes of the forcing ,
the drift is enhanced , but the enhancement is reduced with increasing frequency of the forcing . on the other hand ,
a reduction of the drift or even a flux reversal can be induced at larger amplitudes , while the flux is now found to be an increasing function of the perturbation frequency .
pacs numbers : 02.50.-r , 05.40.+j , 05.60.+w |
peristaltic transport is a natural mechanism of pumping most physiological fluids induced by a progressive wave of area contraction or expansion along the length of the boundary of a fluid - filled distensible tube .
different physiological phenomena , such as the flow of urine from kidney to the bladder through ureters as well as the transport of food material through the digestive tract , semen in the vas differences , fluids in lymphatic vessels , bile from gall bladder into the duodenum , movement of spermatozoa in the ductus efferentes of the male reproductive tract and cervical canal and the transport of ovum in the fallopian tube , the movement of cilia , circulation of blood in small blood vessels , take place by the peristalsis mechanism .
this phenomenon is also observed in the propulsion of some industrial fluids .
[ cols= " < , < " , ] this mechanism finds applications in roller and finger pumps , heart - lung machine , blood pump machine , dialysis machine and also noxious fluid transport in nuclear industries .
some of the recent important theoretical studies on peristaltic transport include those by usha and rao @xcite , misra and pandey @xcite , mishra and rao @xcite , mekheimer @xcite , afifi @xcite , eytan et al .
@xcite , eldabe et al .
@xcite , hakeem et al .
@xcite as well as hayat and ali @xcite .
these studies have been carried out either by using lubrication theory neglecting the fluid inertia and wall curvature without any limitation of wave amplitude or by considering small peristaltic wave amplitude with arbitrary reynolds number . using the assumptions of thin shell and lubrication theories , carew and pedley @xcite put forward a mathematical analysis for the study of peristaltic pumping in the ureter .
antanovskii and ramkissoon @xcite also took the help of lubrication theory in order to study the peristaltic transport of a compressible viscous fluid through a pipe when the pressure drop changes with time , by taking into account the wall deformation of the pipe .
lee and fung @xcite studied theoretically the peristaltic phenomenon observed in blood flow , while lee @xcite reported his experimental studies on peristalsis observed in large intestines .
+ taylor @xcite conducted a theoretical study on asymmetric wave propagation in waving sheets with an object to put forward an explanation to the mechanical interaction between neighbouring spermatozoa . by considering a biharmonic stream function in the form of a fourier series and retaining only the starting term of the stream function series
, he made an attempt to simulate cases for symmetric /asymmetric wave propagation .
the boundary integral method along with a suitable numerical technique was employed by pozrikidis @xcite to study the fluid dynamics involved in the problem of peristaltic transport in an asymmetric channel under stokes flow conditions . a mathematical analysis for peristaltic transport of an incompressible newtonian fluid
was carried out by mishra and rao @xcite as well as elshehaway et al.@xcite . in both these studies ,
channel flow was considered where the channel is asymmetric .
these two studies are similar in nature , however unlike @xcite , the channel was considered to be porous in @xcite .
the work reported in @xcite was extended by reddy et al .
@xcite to the propagation of peristaltic transport of a power - law fluid in an asymmetric channel .
this theoretical study was based on a similar study conducted earlier by reddy et al .
@xcite for the propagation of symmetric peristaltic waves in a newtonian fluid model .
+ recent physiological researches ( cf .
vries et al.@xcite ) indicate that uterine peristalsis resulting from myometrial contraction can take place in both symmetric and asymmetric directions . in different studies pertaining to the gastrointestinal tract ,
intra - pleural membranes , capillary walls , human lung , bile duct , gall bladder with stones and small blood vessels , flow in porous tubes and deformable porous layers have been examined by various authors .
it is known that the primary function of gastrointestinal tract consists of absorbing nutrients from the mix of food and liquid that passes through the tract . as mentioned by keener and sneyd @xcite , gastrointestinal tract
is surrounded by a number of heavily innervated muscle layers which are smooth muscles consisting of many folds . there
exist pores in the junctions between them , although the junctions are tight .
also , bergel @xcite observed that the capillary walls are surrounded by flattened endothelial cell layers which are porous .
+ currently studies on peristaltic motion in magneto hydrodynamic ( mhd ) flows of electrically conducting physiological fluids have become a subject of growing interest for researchers .
this is due to the fact that such studies are useful particularly for having a proper understanding of the functioning of different machines used by clinicians for pumping blood and magnetic resonance imaging ( mri ) .
the related studies also contribute to updating the information that is required to operate machines like mhd peristaltic compressors . as pointed out by agrawal and anwaruddin @xcite , theoretical researches with an aim to explore the effect of a magnetic field on the flow of blood in atherosclerotic vessels also find application in a blood pump used by cardiac surgeons during the surgical procedure .
investigations related to the study of the effect of a magnetic field on the movement of blood indicate that if we apply a frequency pulsing magnetic field of low intensity in a controlled manner , it is possible to modify the behaviour of cells and tissues .
@xcite observed that an impulsive magnetic field can be used as a theraptic means to treat patients who have stone fragments in their urinary tract .
+ in view of all the aforesaid observations and remarks , the present theoretical study has been designed in such a manner that it can explore a variety of information that can enrich the information bank required for magneto therapy , which consists of applying a magnetic field for the treatment of various diseases . from the physiological fluid dynamics point of view , it is a study of the peristaltic transport of a physiological fluid in a porous asymmetric channel under the action of an external magnetic field .
the peristaltic wave train on the channel wall has been considered to have different amplitudes and phases in view of the variation in the channel width , wave amplitude and phase difference .
it is believed that the results presented here will find prospective application in the study of various fluid mechanical problems associated with gastrointestinal tract , intra pleural membranes , capillary walls and small blood vessels .
let us consider the peristaltic motion of an incompressible viscous fluid through a porous medium channel under the action of a magnetic field .
let @xmath0 and @xmath1 be respectively the upper and lower boundaries of the channel .
the medium is considered to be induced by a sinusoidal wave train propagating with a constant speed @xmath2 along the channel wall , such that + @xmath3 on the upper boundary + and + @xmath4 on the lower boundary , + where @xmath5 are the wave amplitudes , @xmath6 is the wave length , @xmath7 the width of the channel , @xmath8 the phase difference ( @xmath9 ) of the waves .
@xmath10 corresponds to the symmetric channel with waves out of phase and @xmath11 represents the waves in phase ; quantities @xmath12 and @xmath8 are such that + @xmath13 + the equations that govern the fluid motion in a porous medium under the influence of a magnetic field are : @xmath14 @xmath15 @xmath16 where @xmath17 , k is the permeability parameter , @xmath18 the intensity of the magnetic field acting along the y - axis and the induced magnetic field is assumed to be negligible . + considering a wave frame @xmath19 moving with a velocity @xmath2 away from a fixed frame @xmath20 , let us the transformations + @xmath21 in which @xmath22 and @xmath23 are the velocity components , @xmath24 and @xmath25 stand for pressure in wave and fixed frame of reference respectively .
+ if @xmath26 be the stream function , we have + @xmath27 .
+ now eliminating p between ( 3 ) and ( 4 ) , we obtain the governing equation expressed in terms of @xmath26 . thus the equation governing the steady flow of the fluid in the wave frame of reference can be written in the form + @xmath28 in the sequal , we shall make use of the following non - dimensional variables : @xmath29 henceforward we shall use the non - dimensional variables defined in ( 7 ) .
if we now drop bars over the symbols , the equation governing the flow of the fluid reads + @xmath30 * boundary conditions : * + considering the physics of the problem , the boundary conditions may be put mathematically as + @xmath31 where @xmath32 refers to the flux in the wave frame and @xmath33 are such that + @xmath34 applying long wave length approximation @xmath35 and assuming the reynolds number to small , the equation ( 8) approximates to @xmath36 where @xmath37 the solution of equation ( 11 ) satisfying ( 9 ) is @xmath38 with @xmath39 @xmath40 @xmath41 @xmath42 where @xmath43 . + in the absence of the magnetic field , these results are found to coincide with those obtained in @xcite , where the porosity effect is disregarded . + the flux at any axial station in the fixed frame is found to be given by @xmath44 while the expression for the volumetric flow rate over one period @xmath45 ) of the peristaltic wave is obtained as @xmath46 the pressure gradient obtained from the dimensionless momentum equation for axial velocity can be expressed as @xmath47 @xmath48 on integration over one wave length it reduces to @xmath49 where @xmath50 @xmath51 in the fixed frame of reference , the expression for the axial velocity is @xmath52 + with @xmath53~~ and~~ h_2=-d - b~cos[2\pi(x - t+\phi)].\ ] ]
analytical expressions for the pressure gradient , the volumetric flow rate , axial velocity and the stream function for the problem under consideration have been calculated and presented in the previous section . in this section , the problem has been investigated further by employing appropriate computational method and numerical techniques . for the purpose of computation , the following set of values ( which are in agreement with experimental data for physiological fluids ) of the different parameters involved in the forgoing theoretical analysis
have been used : + a , b= 0.3 , 0.4 , 0.5 , 0.7 , 1.2 ; d=1.0 , 2.0 ; @xmath54 ; m=0 to 8 ; @xmath55k=0.02 , 0.05 , 0.08 , 0.1 , 0.13 , 0.15 , 0.2 , 0.5 , 1 , 10 , 100 , 1000 ; @xmath56 .
small values of the permeability factor refer to cases when the medium is almost rigid .
+ due to the complexity of the problem under consideration , it was not possible to determine the analytical expression of @xmath57 in terms of @xmath58 and other parameters . thus had to resort to the use of appropriate numerical method along with quadrature formula . similarly , an analytical treatment was found inadequate for finding the velocity as a function of @xmath57 , x , y , t and other parameters that are involved .
this necessitated calculating first + @xmath59 + which was also a very complicated task . here too , we had to make use of numerical quadrature . +
* velocity change * + the distribution of axial velocity in the case of free pumping for different values of the phase difference , the permeability parameter k and the magnetic parameter m is shown in figs 2 - 5 .
fig.2 shows that the results computed on the basis of our study in absence of any external magnetic field ( m=0 ) perfectly match with the results reported in @xcite .
figs.3(a)-(c ) give the changes in the velocity distribution when m=2 , as the phase difference @xmath60 changes from 0 to @xmath61 .
these figures indicate that the velocity changes considerably with change in the values of the amplitudes of the peristaltic wave at the upper and lower walls of the channel .
these figures reveal that the magnitude of the velocity gradually decreases and the extent of flow reversal near the boundary increases as the phase difference increases , when the wave amplitude at the upper boundary of the channel is more than or equal to the wave amplitude at the lower boundary .
it is interesting to note that when the wave amplitude at the lower boundary exceeds that at the upper boundary , the scenario is similar when @xmath60 changes from 0 to @xmath62 . in this case , when @xmath60 exceeds @xmath62 , the flow reversal at the boundary starts decreasing and reduces to zero when @xmath63 .
when @xmath60 increases further , the flow reversal takes place in the central region and the velocity near the boundary increases .
4 shows that flow reversal takes places for all the values of the magnetic parameter examined here and also that while in the central portion of the channel , the axial velocity decreases with the increase of magnetic intensity , the behaviour is slightly opposite in the lower and upper regions .
it may be noted from fig .
5 that in central region of the channel the velocity increases with the increase in the permeability parameter of the channel , while an opposite trend is found in the lower and upper regions .
the occurrence of reverse flow at the initial stage is found to be independent of the permeability factor . + * pumping characteristics * + figs .
6 - 8 illustrate the variation of the volumetric flow rate of peristaltic wave with pressure gradient for different values of the phase difference , permeability factor and the magnetic intensity .
one may observe from fig .
6 that in the range of values of the pressure gradient examined in the present study , in the entire pumping region ( @xmath640 ) , in the free pumping zone ( @xmath650 ) as well as in the co - pumping region ( @xmath660 ) for @xmath64 - 12.3 , the volumetric flow rate decreases with the increase in the phase difference
. however , the trend reverses as soon as the pressure gradient drops below -12.3 .
these observations , of course , hold good for a particular set of values indicated in fig .
6 caption .
7 shows that in the entire pumping region the volumetric flow rate decreases with the increase in permeability parameter , where as in the co - pumping region , a reverse trend is noticed .
one can also observed that the volumetric flow rate can be gradually increased in the pumping region and gradually reduced in the co - pumping region by increasing the intensity of the applied magnetic field . + * trapping * + a theoretical study on peristaltic pumping of a fluid was conducted by shapiro et al .
@xcite for the case when the wave length is large and the reynolds number is small .
this study revealed that at high flow rates and large occlusions , there is a region of closed stream lines in the wave frame , whereby some fluid gets trapped within a wave of propagation .
the trapped fluid mass is found to move with a mean speed equal to that of the wave .
thus the phenomenon of trapping may be looked upon as the formation of an internally circulating bolus of fluid .
owing to the trapping phenomenon , there will exist stagnation points , where both the velocity components of the fluid vanish in the wave frame .
the stagnation points will be located at the points where the central line of the channel intersects the curve @xmath67 ( @xmath26 being the stream function ) . in case of asymmetric channel flow ,
however , some stagnation points may be located elsewhere .
9(a)-(d ) reveals a very interesting fact that with the same amplitude , a = b=0.5 for @xmath54 , the bolus appearing in the central region for @xmath68 moves towards left and decreases in size as @xmath60 increases . for @xmath69 ,
the bolus completely disappears and the stream lines are parallel to one another .
+ the influence of the intensity of the applied magnetic field on the streamlines for peristaltic transport velocity where a = b=0.5 and @xmath55k=1.0 , @xmath60=0.0 and d=1.0 is demonstrated in figs .
10(a)-(d ) .
these figures indicate that when the magnetic field intensity is low , bolus is formed at the central region of the channel .
these bolus is shifted towards the upper boundary of the channel with the increase in the magnetic field strength .
10(d ) reveals that the bolus completely vanishes when m @xmath707.0 .
11(a)-(d ) indicate that when m=1 , in the absence of phase difference , bolus is formed for greater values of the permeability factor but no bolus is formed when the permeability parameter of the channel is small . + a comparison between fig .
11(d ) and fig .
12(a ) reveals that when m=1 , bolus formation is of different type when the phase difference is @xmath71 , even the channel has the same permeability .
but when the intensity of the magnetic field is higher , the bolus disappears even if the phase difference and permeability remain unchanged [ cf . fig .
12(b ) ] . figs .
13(a)-(b ) give us an idea of the type of bolus formation for greater values of the permeability factor .
the variation in the shape of the streamlines and bolus formation for different sets of values of a and b can been seen from figs .
14(a)-(c ) .
the peristaltic transport of a newtonian fluid ( as an idealization of a physiological fluid ) in a porous asymmetric channel under the action of an externally applied magnetic field has been the concern of the present investigation .
the study particularly pertains to the situation when the reynolds number is low and curvature of the channel is quite small .
emphasis has been paid to study some important bio - fluid dynamical phenomena of peristaltic transport such as the pumping characteristics , axial velocity as well as streamline patterns and trapping as a function of asymmetric wave motility parameters , the permeability parameter and the magnetic parameter .
all these have been treated both analytically and numerically .
the results and observations are good agreement with those reported in @xcite .
+ on the basis of the observations of the present study , one can make an important conclusion that it is possible to increase pumping for the adverse pressure gradient by applying an external magnetic field and that the bolus formation can be avoided by suitably adjusting the magnetic field intensity .
the present study reveals further that the velocity change is strongly dependent on the phase difference , amplitudes of wave propagating in the upper and lower regions , magnetic field intensity and the channel permeability .
+ the results throw some light on problems associated with fluid movement in the gastrointestinal tract , intra - uterine fluid motion induced by uterine contraction as well as flow through small blood vessels and intra - pleural membranes .
+ the study suggest that a convenient criterion for the presence of a trapped bolus is the existence of stagnation points in the wave frame .
this physical phenomenon can be considered responsible for formation of thrombus in blood and the movement of food bolus in the gastrointestinal tract . + * acknowledgment : * _ the authors wish to express their deep sense of gratitude to the anonymous referees for their appreciation of the piece of research embodied in the paper , for their words of praise regarding the results reported in the paper and also for their suggestions based upon which the revised manuscript has been prepared .
the authors are also thankful to the council of scientific and industrial research ( csir ) , new delhi for their financial support towards this study .
_ usha , s. and rao , a. r. , `` effect of curvature and inertia on the peristaltic transport in a two fluid system '' , int .
, 38 ( 2000 ) , 1355 - 75 .
misra , j.c . and
pandey , s.k.``peristaltic transport of a non - newtonian fluid with a peripheral layer '' , int . j. engng .
37(1999 ) , 1841 - 58 .
misra , j.c . and pandey , s.k.``peristaltic flow of a multi layered power - law fluid through a cylindrical tube '' , int . j. engng .
39(2001 ) , 387 - 402 .
misra , j.c . and pandey , s.k.``peristaltic transport of a particle - fluid suspension in a cylindrical tube '' , comput . math .
appl.,28(1994 ) , 131 - 45 .
misra , j.c . and pandey , s.k.``peristaltic transport in a tapered tube '' , math . comput
, 22(1995 ) , 137 - 51 .
misra , j.c . and
pandey , s.k.``a mathematical model for oesophegeal swallowing of a food - bolus '' , math .
comput . model .
, 33(2001 ) , 997 - 1009 .
misra , j.c . and
pandey , s.k.``peristaltic transport of blood in small vessels : study of a mathematical model '' , comput . math .
appl.,43(2002 ) , 1183 - 93 . mishra , m. and rao , a.r.``peristaltic transport of a newtonian fluid in an asymmetric channel '' , zamp.54(2003 ) , 532 - 50 .
mishra , m. and rao , a.r.``peristaltic transport of a power law fluid in a porous tube '' , j.non-newtonian fluid mech.121(2004 ) , 163 - 74 .
mekheimer , kh.s .
`` peristaltic flows of blood under effect of a magnetic field in a non uniform channels '' , app.math.comput.153(2004 ) , 763 - 77 .
afifi , n.a.s .
`` interaction of pulsatile and peristaltic transport of a magneto - fluid with suspended particles '' , sci.echoes 10(2006 ) , 9 - 16 .
eytan , o. , jaffa , a.j . and elad , d.``peristaltic flow in a tapered channel : application to embryo transport within the uterine cavity '' , med.engng.phy.23(2001 ) , 473 - 82 .
eldabe , n.t.m , el - sayed , m.f . ,
ghaly , a.y . and
sayed , h.m.``peristically induced transport of a mhd biviscosity fluid in a non - uniform tube '' , physica a 383(2007 ) , 253 - 66 .
hakeem , a.el . , naby , a.el . ,
misiery , a.e.m.el . and shamy , i.el.``hydromagnetic flow of generalized newtonian fluid through a uniform tube with peristalsis '' , app.math.comput .
173(2006 ) , 856 - 71 .
hayat , t. and ali , n. `` peristaltic motion of a jeffrey fluid under the effect of a magnetic field in a tube '' , commun.nonlinear sci.num.simu.13(7 ) ( 2008 ) , 1343 - 1352 .
carew , e.o and pedley , t.j `` an active membrane model for peristaltic pumping : part i- periodic activation waves in an infinite tube '' , j. biomech . eng . , 119(1997 ) , 66 - 76 .
antanovskii , l.k and ramkissoon , h `` long - wave peristaltic transport of a compressible viscous fluid in a finite pipe subject to a time - dependent pressure drop '' , fluid dynamics research .
19(1997 ) , 115 - 123 .
lee , j.s . ,
fung , y.c .
flow in nonuniform small blood vessels
, microvasc .
, 3(1971 ) , 272 - 87 .
lee , c.y `` the effect of stimulation of extrinsic nerves on peristalsis and on the release of 5-hydroxytryptamine in the large intestine of the guinea - pig and of the rabbit '' , j physiol 152(1960 ) , 405 - 18 .
lee , c.y `` a study of peristalsis of the large intestine in vitro '' , j physiol 148(1959 ) , 63 - 64 .
taylor , g.i .
analysis of the swimming of microscopic organisms ,
proc.roy.soc.lond . a 209 ( 1951 ) , 447 - 61 .
pozrikidis , c.a study of peristaltic flow , j.fluid mech .
180 ( 1987 ) , 515 - 27 .
elshehawey , e.f , eldabe , n.t elghazy , e.m and ebaid , a. peristaltic transport in an asymmetric channel through a porous medium ,
app.math.comput.182(2006 ) , 140 - 50 . reddy , m.v.s . , rao , a.r and sreenadh , s.peristaltic motion of a power - law fluid in an asymmetric channel , int .
j. non - linear mech . 42 ( 2007 ) , 1153 - 61 .
reddy , m.v.s . , mishra , m. , sreenadh , s. and rao , a.r peristaltic motion of a power - law fluid in an asymmetric channel,j.fluids engng : .
asme 127(2005 ) , 824 - 27 .
vries , k.d . , lyons , e.a . , ballard , j. , levi , c.s . and
lindsay , d.j .
contractions of the inner third of myometrium , am.j.obstetries gynecol . 162 ( 1990 ) , 679 - 82 .
keener , j.r . and sneyd , j. mathematical physiology , springer , 1998 .
berrgel , d.h .
cardiovascular fluid dynamics , academic press , london , 1972 .
agrabal , h.l . and anwaruddin .
b.``peristaltic flow of blood in a branch '' , ranchi univ.math.j.15(1984 ) , 111 - 21 .
li , a. , nesterov , n.i .
malikova , s.n . and
kiiatkin , v.a.``the use of an impulse magnetic field in the combined of patients with stone fragments in the upper urinary tract '' , vopr kurortol fizioter lech fiz kult 3(1994 ) , 22 - 24 .
shapiro , a.h . ,
jaffrin , m.y . and weinberg , s.l .
`` peristaltic pumping with long wavelength at low reynolds number '' , j. fluid mech .
37(1969 ) , 799 - 825 .
mekheimer , kh.s . and al - arabi , t.h.``nonlinear peristaltic transport of mhd flow through a porous medium '' , int.j.math.math.sci.26(2003 ) , 1663 - 82 .
3a : distribution of axial velocity for different values of @xmath8 , where x=1 , t=1 , @xmath57=0 , a=0.5 , b=0.3 , d=1 , @xmath55k=0.8 , m=2 + + fig .
3b : distribution of axial velocity for different values of @xmath8 , where x=1 , t=1 , @xmath57=0 , a=0.5 , b=0.5 , d=1 , @xmath55k=0.8 , m=2 + + fig .
3c : distribution of axial velocity for different values of @xmath8 , where x=1 , t=1 , @xmath57=0 , a=0.3 , b=0.5 , d=1 , @xmath55k=0.8 , m=2 + + fig .
4 : distribution of axial velocity for different values of m , where x=1 , t=1 , @xmath57=0 , a=0.5 , b=0.3 , d=1 , @xmath55k=0.8 , @xmath8=@xmath71 + + fig . 5 : distribution of axial velocity for different values of @xmath55k , where x=1 , t=1 , @xmath57=0 , a=0.5 , b=0.3 , d=1 , m=2 , @xmath8=@xmath71 + + fig . 6 : variation of @xmath58 with @xmath57 for different values of @xmath8 along with a=0.7 , b=1.2 , d=2 , @xmath55k=0.1 , m=2 + + fig . 7 : variation of @xmath58 with @xmath57 for different values of @xmath55k along with a=0.7 , b=1.2 , d=2 , m=2 , @xmath8=@xmath71 + + fig . 8 : variation of @xmath58 with @xmath57 for different values of m along with a=0.7 , b=1.2 , d=2 , @xmath55k=0.1 , @xmath8=@xmath71 + | the paper deals with a theoretical investigation of the peristaltic transport of a physiological fluid in a porous asymmetric channel under the action of a magnetic field . the stream function , pressure gradient and axial velocity are studied by using appropriate analytical and numerical techniques .
effects of different physical parameters such as permeability , phase difference , wave amplitude and magnetic parameter on the velocity , pumping characteristics , streamline pattern and trapping are investigated with particular emphasis .
the computational results are presented in graphical form .
the results are found to be in perfect agreement with those of a previous study carried out for a non - porous channel in the absence of a magnetic field .
+ _ keywords : peristaltic transport , asymmetric channel , flow reversal , trapping . _ 0.8 cm -.1 in -.1 in 0.0 cm 22.5 cm [ section ] [ theorem]definition [ theorem]observation [ theorem ] [ theorem]proposition [ theorem]rule 22.5 cm = currsize |
ever since pauli postulated the neutrino , its influence and significance for particle physics as well as for astrophysics has steadily been growing .
the observation of flavour oscillations of atmospheric and solar neutrinos , as well as of reactor and accelerator neutrinos at long baseline , is a compelling evidence that neutrinos are massive .
due to this fact and owing to their large abundance in the universe , neutrinos are considered as the primary candidate for hot dark matter in cosmology .
thus , they could play an important role for the evolution of large scale structures ( lss ) in the universe . on the other hand ,
the on - going investigations of neutrino properties , their mass hierarchy and especially their rest mass will open a door to the understanding of the origin of mass . here , a precision measurement of the neutrino rest mass can discriminate between different @xmath3-mass models , in particular whether they are of hierarchical type @xmath4 or of quasi - degenerate type @xmath5 . although experiments on neutrino flavour oscillation provide compelling evidence that neutrinos are massive , they can not provide an absolute mass value .
therefore , other methods have to be applied .
cosmological investigations provide very sensitive methods to determine an estimate for the neutrino mass . here , the sum of all neutrino mass eigenstates is being measured , since in cosmological processes the flavour can not be distinguished . using only wmap7 data , an upper limit of 1.2 ev@xmath6
is determined for the sum of neutrino mass eigenstates @xcite .
upcoming data from the planck satellite as well as from lsst and weak lensing surveys have the potential of probing the sum of neutrino masses with 50 mev@xmath6 sensitivity .
however , while cosmologcal studies offer a very sensitive approach , one has to emphasize its generic model dependence , which easily leads to a factor of 2 of uncertainty in the neutrino mass prediction @xcite .
therefore , it is essential to probe the neutrino mass with sub - ev sensitivity in laboratory experiments .
in general , there are two approaches which are complementary in their physics objectives .
one is the search for neutrinoless double @xmath0-decay ( @xmath7 ) , the other is the precise spectroscopy of @xmath0-decay at its kinematic end - point .
the @xmath7 process requires the neutrino to be a majorana particle , thus it appears identical to its anti - particle . from neutrinoless double @xmath0-decay an effective majorana mass @xmath8
is determined , which corresponds to the coherent sum of all mass eigenstates @xmath9 with respect to the pmns mixing matrix @xmath10 .
@xmath11 here , not @xmath12 but @xmath13 appears , hence @xmath14 depends on complex cp - phases with the possibility of cancellations .
this generic fact has to be taken into account in terms of a model - dependence when comparing claims or limits like @xmath15 @xcite to results from single @xmath0-decay experiments .
nevertheless , upcoming @xmath7 experiments like gerda , majorana , sno+ , exo , and cuore have the potential to probe @xmath8 in the 20 - 50 mev/@xmath16 region . in @xcite , and @xcite the experimental and theoretical status of this field as well as the influences by matrix elements are discussed in more detail .
experiments investigating single @xmath0-decay offer a direct and model - independent method to determine the absolute neutrino mass , since they rely only on the relativistic energy - momentum relation and energy conservation @xcite . in single @xmath0-decay the squared neutrino mass @xmath17
is determined as an incoherent sum in contrast to ( [ doppelbeta ] ) of all mass eigenstates according to the pmns matrix .
@xmath18 currently , the best experimental bounds from single @xmath0-decay ( @xmath19 ) have been determined in the tritium @xmath0-decay experiments in mainz @xcite and troitzk @xcite , which have reached their sensitivity limit . at present new experiments
are being assembled or designed which will increase the experimental precision by two orders of magnitude , thus increasing the sensitivity on @xmath20 by one order of magnitude to 200 mev / c@xmath21 . the basic principle applied in this model - independent method
is based on kinematics and energy conservation only , with the squared neutrino mass eigenstate @xmath22 appearing in the phase space factor . for a detailed description including
final states and relativistic kinematics refer to @xcite .
@xmath23 since the experimental resolution at present does not allow to resolve individual mass eigenstates , only a combined spectrum can be recorded , according to @xmath24 the electron neutrino mass parameter corresponding to ( [ singlebeta ] ) is derived by analyzing the spectrum close to the end - point , where @xmath25 is small and the mass term @xmath26 becomes significant .
a non - zero neutrino mass will not only shift the end - point , but also change the spectral shape .
the count rate at the end - point is very low and the absolute q - value is not known precisely enough , therefore it is far more important to measure the shape of the spectrum , rather than the end - point shift .
a low end - point @xmath0 source for a large fraction of electrons in the end - point region is a key requirement as well as a high source luminosity and high energy resolution in addition to very low background . for single @xmath0-decay investigations
there are two complementary approaches with different systematics . 1 . the calorimeter approach , where the source is identical to the detector .
+ here , metallic or dielectric @xmath27re crystal bolometers are used to capture the entire @xmath0-decay energy as differential energy spectrum . with @xmath28
kev @xmath27re has the lowest end - point , but due to its rather long half - life of @xmath29 y the activity is low . since bolometers are modular , their number can be scaled in order to increase the sensitivity .
this approach is being followed in the mare experiment .
the spectrometer approach , where an external tritium source is used .
+ in this case , the kinetic energy of the @xmath0-electrons is analyzed as integral spectrum by an electrostatic spectrometer . here
, the source material is high purity molecular tritium @xmath30 with a low end - point at @xmath31 kev and a short half - life of @xmath32 , providing a high activity .
this approach implies strict scaling laws for the size of the spectrometer and reaches its ultimate size and precision in the katrin experiment . besides these approaches new ideas have recently come up .
the project 8 proposal @xcite aims to make use of radio - frequency spectroscopy in order to measure the kinetic energy of @xmath0 electrons from a gaseous @xmath30 source . here
, an array of antennas would capture the coherent cyclotron radiation emitted by the @xmath0-decay electrons when moving through a homogenous magnetic field . in @xcite
the authors estimate an ultimate sensitivity of @xmath33 ev for this method . currently , a proof - of - principle experiment is in preparation in order to show that it is feasible to detect electrons and to determine their kinetic energy with this method . for details
please see @xcite .
as outlined above , mare ( microcalorimeter arrays for a rhenium experiment ) uses @xmath27re as @xmath0 emitter .
the long half - life of this isotope is the result of the first order unique forbidden transition @xmath34 .
mare will thus be performed in two phases with increased sensitivity .
it is the successor of the manu and mibeta experiments , which derived an upper limit of @xmath35 ev from @xmath36 @xmath0-decays .
phase i of mare aims to improve the current sensitivity on @xmath37 by one order of magnitude by increasing the statistics to @xmath38 @xmath0-decays .
this will result in a sensitivity comparable to the mainz and troitzk experiments ( @xmath39 ev ) , thus scrutinizing their upper limit .
using calorimeter pixel arrays , it will take three years to reach the intended number of decays .
this time is used in parallel to finish the r&d program in order to improve the sensor technology for the next phase ii .
the goal of mare phase ii is to improve the sensitivity by another order of magnitude with increased statistics of @xmath40 @xmath0-decays . with a sensitivity of @xmath41 ev mare will then be in a position to scrutinize a katrin result with a completely independent method and different systematics . in order to reach the projected sensitivity the operation of 50000 bolometer pixels
is required over a measurement time of at least 5 years .
detailed information about the present status and the progress of both mare phases can be found in @xcite , in addition to a review of the investigation of @xmath42ho as alternative @xmath0 emitter .
the katrin ( karlsruhe tritium neutrino ) experiment @xcite makes use of a molecular gaseous tritium source and an electrostatic energy filter based on the mac - e filter principle @xcite .
an overview of the main beam - line is given in figure [ fig : katrin ] .
the mac - e filter uses magnetic adiabatic collimation in combination with an electrostatic energy filter .
electrons emitted in the source at high magnetic field enter the spectrometer , where the magnetic field drops by several orders of magnitude . maintaining a strictly adiabatic regime with full energy conservation
, the magnetic gradient force transforms the cyclotron energy of the isotropically emitted electrons into the longitudinal component . at the same time
, the increasing potential of the electrostatic filter slows down the electrons by reducing their longitudinal energy .
both effects have to be aligned very carefully for a proper energy analysis .
only electrons which are able to cross the potential in the spectrometer are counted , thus the mac - e filter acts as high - pass filter for electrons .
its relative resolution is defined by the magnetic field ratio between minimum field in the center of the spectrometer and maximum field at the pinch magnet .
this defines the overall electromagnetic design of katrin and the size of its components .
katrin faces challenges in all key technologies which are applied in the experiment . as an example
, about @xmath43 electrons / s are emitted in the source , while at the far detector side a background rate of the order of @xmath44 cps has to be reached .
the source column density , temperature and pressure has to be kept stable at the @xmath45 level .
the transport section has to reduce the tritium gas flow by a huge amount of @xmath40 , in order not to increase the background in the spectrometer section .
in addition the decay electrons have to be transported fully adiabatically on the mev - scale over a distance of about 50 m. in the spectrometer region , carefully designed electromagnetic field conditions have to be implemented with high voltage stability on the @xmath46 scale for operating voltages up to @xmath47 kv . in particular , electromagnetic conditions leading to penning - like trapping conditions have to be avoided in order to reach low background rates .
the following subsection addresses the details of some key components and their present status .
the katrin setup stretches over 70 m. the windowless gaseous tritium source ( wgts ) provides a high activity of @xmath43 @xmath0-decays per second .
the tritium gas is injected in the center part of the source beam - tube and flows to both ends , where it is removed again by pumps .
electrons emitted by the t@xmath48-decay are guided by strong magnetic fields ( @xmath49 t in the source , @xmath50 t in the transport section ) to the transport section and finally to the spectrometer section , while at the same time retaining the gas flow by 14 orders of magnitude .
the pre - spectrometer which is a mac - e filter operating at moderate resolution of about @xmath51 ev can be used to transmit only electrons with energies close to the t@xmath48 end - point . in this operation mode
, only electrons of the end - point region would enter the main spectrometer for the precise energy analysis .
the main spectrometer is a mac - e filter and offers a resolution of @xmath52 ev resolution for @xmath31 kev electrons by applying a magnetic field ratio of @xmath53 .
the latter defines the size of the main spectrometer , since the full magnetic flux tube of the source has to be analyzed . in order to meet its sensitivity goal
, katrin requires a stable source with high luminosity , and low systematic effects .
the design luminosity of @xmath54 bq requires @xmath55 t@xmath48 molecules to be injected per second into the source tube , which adds up to a throughput of 40 g per day .
a unique research facility in europe providing the corresponding tritium technology is tritium laboratory karlsruhe ( tlk ) , located at the kit campus north site .
the tritium purity has to be kept above 95% , the source temperature at @xmath56 k , and the injection pressure at @xmath45 mbar . in order to achieve a stable column density all these parameters have to be stable on the @xmath45 level . at tlk
a sophisticated @xmath30 loop system has been commissioned for katrin , providing a stable t@xmath48 injection rate and a high purity isotope separation stage @xcite .
being designed for a stability of @xmath45 , test runs showed that the loop system reaches a @xmath57 stability level for over 4 months .
a new concept of temperature stabilization is applied to the wgts beam tube ( @xmath58 m , @xmath59
cm ) in order to achieve a stability of @xmath60 at the source operating temperature .
the inner beam - tube is directly coupled to a two phase ne thermosiphon @xcite .
this novel concept is presently being tested at the wgts demonstrator , which is the inner part of the wgts cryostat without superconducting magnets .
after the demonstrator tests the wgts cryostat will be completed in order to be ready for implementation into the katrin beam - line in 2012 .
adjacent to the wgts the transport section guides the decay electrons to the spectrometer section by strong magnetic fields of @xmath50 t. tritium retention by the factor @xmath40 is achieved in two steps .
first the differential pumping section dps uses active pumping to provide a t@xmath48 retention of @xmath61 .
the dps2-f cryostat is on - site and its commissioning has recently been completed .
the cryogenic pumping section cps uses cryo - sorption to achieve a further retention of @xmath62 . here
, the liquid he cold inner beam - tube is covered with ar frost , providing a large surface to cryo - sorp the t@xmath48 .
the cps concept has been successfully tested in @xcite , the cryostat is expected to be on - site in mid 2011 .
the spectrometer section consists of two mac - e filters .
the pre - spectrometer is a mac - e filter with sufficient resolution to cut the low energy part of the @xmath0 spectrum at about 300 ev below the end - point , if required . at present
it acts as an important test setup , where all major technologies for the main spectrometer have been tested , i.e. the vacuum concept , the electromagnetic design as well as electron transmission properties and the detector concept .
especially the electromagnetic design will have a strong impact on the background reduction .
even tiny @xmath63-sized penning - like traps close to the ground electrode can cause background rates of @xmath64 cps .
based on the advanced katrin field calculation code , the electrode shape has been refined avoiding these traps and resulting in a background level of a few @xmath45 cps .
this rate can not be distinguished from the detector background , hence the pre - spectrometer is considered as quasi background free .
remaining background features have originated from rn decays in the volume , which come from auxiliary components and the seas non - evaporable getter material .
the auxiliary parts involved have been removed and rn from the getter material has been eliminated by ln@xmath48 baffles , for details see @xcite .
the pre - spectrometer will be integrated into the beam - line in 2011 .
the main spectrometer analyses the @xmath0-decay electrons of the end - point region with a resolution of @xmath52 ev and has to operate with a background rate at the @xmath45 cps level .
therefore , background from low - energy electrons created by cosmic muons in the vessel wall have to be rejected . for this purpose not only a magnetic shielding is being relied upon , but also a uhv compatible wire electrode system ( 240 modules , 23000 wires , @xmath65 mm precision ) is currently being installed , see figure [ fig : mainspec ] .
vacuum tank and wire electrode provide the precise retarding potential for the energy filter and are supplied by high - voltage equipment at a precision level of @xmath46 .
this level is of key importance for the @xmath3-mass measurement , thus it will be monitored by ultra - precise high - voltage dividers @xcite and a monitor spectrometer beam - line .
the latter uses the same retarding potential and measures mono - energetic electron sources .
since the electromagnetic properties of the mac - e filter are extremely sensitive to distortions of the magnetic field , an air coil system has been installed .
as shown in figure [ fig : mainspec ] it consists of large coils around the vacuum vessel .
the air coil system compensates the earth magnetic field and in addition it is used for the fine tuning of the magnetic flux tube .
the transmitted electrons are counted by a detector based on a monolithic 148 pixel si - pin diode with an energy resolution of about 1 kev , and the ability to detect rates from @xmath45 cps in the end - point region and up to @xmath66 cps during calibration runs , while keeping the background low .
mechanically it is designed like a dartboard with 9 cm diameter and 12 concentric rings with @xmath67 segmentation .
the detector setup including magnets , vacuum components and daq is being delivered for system integration in early 2011 @xcite .
after commissioning and testing , the complete system integration is planned for late 2012 . as soon as the setup is in measurement condition , katrin will take 3 years of data , which corresponds to 5 years real - time .
the statistical uncertainty is estimated to be @xmath68 ev@xmath69 .
likewise , the systematic uncertainty is being restricted to @xmath70 ev@xmath69 in order to minimize the total uncertainty to @xmath71 ev@xmath69 .
this leads to the katrin design sensitivity of @xmath72 ev@xmath6 ( 90% cl ) and the discovery potential of @xmath73 ev@xmath6 with 5@xmath74 significance .
the on - going efforts to measure the absolute mass of neutrinos is well motivated by recent results in particle and astroparticle physics .
the investigation of @xmath0-decay offers a model - independent method for determining the @xmath3-mass .
the re - based mare experiment is going to check the present limits , while at the same time improving the detector technology . in future
, mare will be able to scrutinize a katrin result , after operating 50000 bolometer pixels for at least 5 years .
new concepts for @xmath0-spectroscopy such as projects 8 will open new ways to optain sub - ev sensitivities and should be matured in an r&d phase .
the estimated sensitivity of the tritium - based katrin experiment is @xmath72 ev/@xmath16 ( 90% cl ) .
its construction is proceeding well with several of the major components being already on - site .
the main spectrometer test program will start in 2011 and the complete system integration is planned for 2012 .
00 s. hannestad _ et al . _ , arxiv:1004.0695 m. c. gonzalez - garcia _ et al .
_ , jhep * 08 * ( 2010 ) 117 , arxiv:1006.3795 h. v. klapdor - kleingrothaus _ et al .
_ , eur . phys .
j. a * 12 * ( 2001 ) 147 w. rodejohann , `` neutrinoless double beta decay in particle physics '' , these proceedings m. pavan , `` introduction to double beta decay experiments and cuore '' , these proceedings m. dolinski , `` the enriched xenon observatory ( exo ) '' , these proceedings k. nakamura , `` scintillator - based experiments in double beta decay '' , these proceedings f. simkovic , `` matrix elements for double beta decay '' , these proceedings s. s. masood _ et al .
_ , phys .
c * 76 * ( 2007 ) 045501 e. w. otten and ch .
weinheimer , rep .
prog . phys .
* 71 * ( 2008 ) 086201 c. kraus _ et al .
_ , eur .
j. c. * 40 * ( 2005 ) 447 v. m. lobashev , nucl .
a719 * ( 2003 ) 153c b. monreal _ et al . _ , phys . rev .
d * 80 * ( 2009 ) 051301(r ) j. a. formaggio , `` project 8 : measuring neutrino mass using radio - frequency techniques '' , these proceedings a. nucciotti , `` neutrino mass calorimetric searches in the mare experiment '' , these proceedings a. osipowicz _ et al .
_ , ( katrin collaboration ) , arxiv : hep - ex/0109033 ( 2001 ) j. angrik _ et al . _ , fzk scientific report 7090 , bibliothek.fzk.de/zb/berichte/fzka7090.pdf a. picard _ et al . _ , nucl . instr
* 63 * ( 1992 ) 345 s. fischer , `` laser raman spectroscopy for katrin '' , these proceedings s. grohmann , cryogenics , volume 49 , issue 8 ( 2009 ) 413 - 420 o. kazachenko _ et al
. _ , nuclear instruments and methods in physics research a * 587 * ( 2008 ) 136 f. eichelhardt _ et al
. _ , fusion science and technology * 54 * ( 2008 ) 615 f. m. frnkle , dissertation ( 2010 ) , digbib.ubka.uni-karlsruhe.de/volltexte/1000019392 r. marx , ieee trans . on instr . and meas .
50 no . 2 ( 2001 ) 426 - 429 t. thmmler
_ , new j. phys .
* 11 * ( 2009 ) 103007 b. wall , `` overview of the katrin detector section '' , these proceedings p. renschler , `` mc simulations of the detector response to low energy electrons for katrin '' , these proceedings | the properties of neutrinos and especially their rest mass play an important role at the intersections of cosmology , particle physics and astroparticle physics . at present
there are two complementary approaches to address this topic in laboratory experiments .
the search for neutrinoless double beta decay probes whether neutrinos are majorana particles and determines an effective neutrino mass value . on the other hand experiments such as mare , katrin and
the recently proposed project 8 will investigate the spectral shape of @xmath0-decay electrons close to their kinematic endpoint in order to determine the neutrino rest mass with a model - independent method . here ,
because of neutrino flavour mixing , the neutrino mass appears as an average of all neutrino mass eigenstates contributing to the electron neutrino . the karlsruhe tritium neutrino experiment ( katrin )
is currently the experiment in the most advanced status of commissioning .
it combines an ultra - luminous molecular windowless gaseous tritium source with an integrating high - resolution spectrometer of mac - e filter type . it will investigate the neutrino rest mass with @xmath1 ( 90% c.l . )
sensitivity and allow @xmath0 spectroscopy close to the @xmath2 endpoint at 18.6 kev with unprecedented precision .
direct neutrino mass , beta decay spectroscopy , mare , katrin , project 8 |
Subsets and Splits